04_Densoviruses/04_Lab_journal.ipynb

{
 "cells": [
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "# **CHAPTER 4. Densoviruses**"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "**Install conda env and activate it**"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "```\n",
    "conda env create -f denso.yaml\n",
    "```\n",
    "\n",
    "```\n",
    "conda activate denso\n",
    "```"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## **Part 1. Prepare to extract densoviruses from MAGs**"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "Create a directory where complete genomes of densoviruses from `RefSeq` database will be stored"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 1,
   "metadata": {},
   "outputs": [],
   "source": [
    "! mkdir genbank_seqs"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "Download densoviruses complete genomes from `RefSeq` database to `genbank_seqs/refseq_densoviruses.fa` file"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 4,
   "metadata": {},
   "outputs": [],
   "source": [
    "! esearch -email iljapopov17@gmail.com -db nucleotide -query 'Densovirus AND \"complete genome\" AND srcdb_refseq[PROP]' | efetch -format fasta -mode text > genbank_seqs/densoviruses_genomes.fa"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "Create a directory where `BLAST` custom database of densoviruses complete genomes will be stored"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 8,
   "metadata": {},
   "outputs": [],
   "source": [
    "! mkdir densovirus_blast_db"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "Create `BLAST` database"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 9,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "\n",
      "\n",
      "Building a new DB, current time: 08/06/2024 18:02:19\n",
      "New DB name:   /home/user/BatMetagenome2x150/Clean_journals/Densoviruses/densovirus_custom_db/densovirus_db\n",
      "New DB title:  refseq_seqs/refseq_densoviruses.fa\n",
      "Sequence type: Nucleotide\n",
      "Keep MBits: T\n",
      "Maximum file size: 3000000000B\n",
      "Adding sequences from FASTA; added 56 sequences in 0.022052 seconds.\n",
      "\n",
      "\n"
     ]
    }
   ],
   "source": [
    "! makeblastdb -in refseq_seqs/refseq_densoviruses.fa -dbtype nucl -out densovirus_blast_db/densovirus_db"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## **Part 2. Extract densoviruses from MAGs**"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    ">Disclaimer: it is important to install `VirSorter2` with all the requirements and dependecies. <br>\n",
    "> Although `VirSorter2` will be installed with `denso.yaml` conda envinronment it is also important to install the database needed. Please follow the official [installation guide](https://github.com/jiarong/VirSorter2#download-database-and-dependencies)<br>\n",
    "> In the `denso.yaml` `VirSorter2` is already installed with `conda` so proceed to:\n",
    "> ```\n",
    "># just in case there is a failed attemp before; \n",
    ">#   remove the whole diretory specified by -d\n",
    ">rm -rf virsorter_db\n",
    "># run setup\n",
    ">virsorter setup -d virsorter_db -j 4\n",
    "> ```"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "`VirExtr_pipeline` is a Snakemake pipeline that runs filtered MAGs from the previous chapter against custom `BLAST` database with reference genomes of densoviruses. Then it extracts longest alignment sequences to a `.fasta` files. Finally, it runs `VirSorter2` to assess viral genomes."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [],
   "source": [
    "! snakemake -s VirExtr_pipeline --cores all --use-conda"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "As the result of this pipeline:\n",
    "- `BLAST` results are stored in `BLAST_vs_densovirus` folder<br>\n",
    "- Extracted densoviruses are stored in `viral_genomes` folder<br>\n",
    "- `VirSorter2` results are sorted in `VirSorted2_results` folder"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "Extracted densoviruses genomes are named like this:<br>\n",
    "`>NODE_632_length_4705_cov_657475.178280`<br>\n",
    "It is no use! `Prokka` will not take it as an input as the seqs names are too long<br>\n",
    "\n",
    "That is why we will rename sequences:<br>\n",
    "- Densoviruses from `D` samples will be renamed to `Densovirinae_sp._isolate_RBHC_VM_#`<br>\n",
    "- Densoviruses from `P` samples will be renamed to `Densovirinae_sp._isolate_RBHC_NN_#`<br>\n",
    "\n",
    "`VM` stands for _Vespertilio murinus_<br>\n",
    "`NN` stands for _Nyctalus noctula_"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 7,
   "metadata": {},
   "outputs": [],
   "source": [
    "# Usage\n",
    "# {path_to_script} {path_to_input_folder}\n",
    "%run scripts/rename_seqs.py viral_genomes/"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "Let's check how the script worked"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 1,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      ">Densovirinae_sp._isolate_RBHC_VM_1\n"
     ]
    }
   ],
   "source": [
    "! head -1 viral_genomes/Densovirus_D1.fasta"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "That's it!"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## **Part 3. Prepare for genome annotation of densoviruses**"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "Download densoviruses proteins from `Protein` database to `genbank_seqs/densoviruses_proteins.fa` file"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 5,
   "metadata": {},
   "outputs": [],
   "source": [
    "! esearch -email iljapopov17@gmail.com -db protein -query 'Densovirus' | efetch -format fasta -mode text > genbank_seqs/densoviruses_proteins.fa"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## **Part 4. Genome annotation of densoviruses**"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "### **Part 4.1. Genome annotation**"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "`GA_pipeline` is a Snakemake pipeline that annotates isolated from MAGs densoviral genomes with `prokka` and densoviruses proteins from NCBI `Proteins` database downloaded in the cell above."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [],
   "source": [
    "! snakemake -s GA_pipeline --cores all"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "As the result of this pipeline:\n",
    "- `prokka` results are stored in `prokka_results` folder"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "### **Part 4.2. Fix `prokka` `.gbk` outputs**"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "Let's look at the `.gbk` files, especially at the 1st row of the file"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 8,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "LOCUS       Densovirinae_sp._isolate_RBHC_VM_14705 bp   DNA linear\n"
     ]
    }
   ],
   "source": [
    "! head -1 prokka_results/annotated_densovirus_D1/Densovirus_D1_annot.gbk"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "Well... `prokka` makes a headache problem...<br>\n",
    "There is no space between sample number and seq length! `Proksee` will not take this `.gbk` as an input!<br>\n",
    "Let's fix it!"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "`scripts/add_space_gbk.py` will fix \"space problem\". Just run this script. It will make changes exactly in the same file."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 6,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Processing file: prokka_results/annotated_densovirus_D1/Densovirus_D1_annot.gbk\n",
      "Processing file: prokka_results/annotated_densovirus_D2/Densovirus_D2_annot.gbk\n",
      "Processing file: prokka_results/annotated_densovirus_D3/Densovirus_D3_annot.gbk\n",
      "Processing file: prokka_results/annotated_densovirus_D4/Densovirus_D4_annot.gbk\n",
      "Processing file: prokka_results/annotated_densovirus_D5/Densovirus_D5_annot.gbk\n",
      "Processing file: prokka_results/annotated_densovirus_P1/Densovirus_P1_annot.gbk\n",
      "Processing file: prokka_results/annotated_densovirus_P2/Densovirus_P2_annot.gbk\n",
      "Processing file: prokka_results/annotated_densovirus_P3/Densovirus_P3_annot.gbk\n",
      "Processing file: prokka_results/annotated_densovirus_P4/Densovirus_P4_annot.gbk\n",
      "Processing file: prokka_results/annotated_densovirus_P5/Densovirus_P5_annot.gbk\n"
     ]
    }
   ],
   "source": [
    "# Usage\n",
    "# {path_to_script}\n",
    "%run scripts/add_space_gbk.py"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 9,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "LOCUS       Densovirinae_sp._isolate_RBHC_VM_1 4705 bp   DNA linear\n"
     ]
    }
   ],
   "source": [
    "! head -1 prokka_results/annotated_densovirus_D1/Densovirus_D1_annot.gbk"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "Hooray!"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "### **Part 4.3. Visualize**"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "Please visit [`Proksee`](https://proksee.ca/) website and upload `.gbk` files from `prokka_results/annotated_densovirus_{sample}` folders."
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## **Part 5. Comparison of densoviruses**"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "### **Part 5.1. Install `fastANI`**"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "Download it from `GitHub`"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [],
   "source": [
    "! wget https://github.com/ParBLiSS/FastANI/releases/download/v1.34/fastANI-linux64-v1.34.zip"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "Make directory to store `ANI` files and results"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [],
   "source": [
    "! mkdir ANI"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "Unzip donwloaded from `GitHub` archive and delete it"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [],
   "source": [
    "! unzip fastANI-linux64-v1.34.zip -d ANI && rm -rf fastANI-linux64-v1.34.zip"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "### **Part 5.2. Run `fastANI`**"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [],
   "source": [
    "! ANI/fastANI --ql ANI/querylist.txt --rl ANI/reflist.txt -o ANI/fastani.out --matrix"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "### **Part 5.3. Visualize**"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "Rename `viral_genomes/Densovirus_{sample}.fasta` to `Densovirinae_sp._isolate_RBHC_{sample}`"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 80,
   "metadata": {},
   "outputs": [],
   "source": [
    "# Load the contents of the file\n",
    "file_path = 'ANI/fastani.out'\n",
    "\n",
    "with open(file_path, 'r') as file:\n",
    "    file_content = file.read()\n",
    "\n",
    "# Define the mapping for renaming the files\n",
    "rename_map = {\n",
    "    \"viral_genomes/Densovirus_D1.fasta\": \"Densovirinae_sp._isolate_RBHC_VM_1\",\n",
    "    \"viral_genomes/Densovirus_D2.fasta\": \"Densovirinae_sp._isolate_RBHC_VM_2\",\n",
    "    \"viral_genomes/Densovirus_D3.fasta\": \"Densovirinae_sp._isolate_RBHC_VM_3\",\n",
    "    \"viral_genomes/Densovirus_D4.fasta\": \"Densovirinae_sp._isolate_RBHC_VM_4\",\n",
    "    \"viral_genomes/Densovirus_D5.fasta\": \"Densovirinae_sp._isolate_RBHC_VM_5\",\n",
    "    \"viral_genomes/Densovirus_P1.fasta\": \"Densovirinae_sp._isolate_RBHC_NN_1\",\n",
    "    \"viral_genomes/Densovirus_P2.fasta\": \"Densovirinae_sp._isolate_RBHC_NN_2\",\n",
    "    \"viral_genomes/Densovirus_P3.fasta\": \"Densovirinae_sp._isolate_RBHC_NN_3\",\n",
    "    \"viral_genomes/Densovirus_P4.fasta\": \"Densovirinae_sp._isolate_RBHC_NN_4\",\n",
    "    \"viral_genomes/Densovirus_P5.fasta\": \"Densovirinae_sp._isolate_RBHC_NN_5\",\n",
    "}\n",
    "\n",
    "# Replace the names in the file content\n",
    "for old_name, new_name in rename_map.items():\n",
    "    file_content = file_content.replace(old_name, new_name)\n",
    "\n",
    "# Save the updated content back to the file\n",
    "with open(file_path, 'w') as file:\n",
    "    file.write(file_content)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "Import needed `python` modules"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 138,
   "metadata": {},
   "outputs": [],
   "source": [
    "import pandas as pd\n",
    "import seaborn as sns\n",
    "import matplotlib.pyplot as plt"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "Make directory to store output image"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 86,
   "metadata": {},
   "outputs": [],
   "source": [
    "! mkdir images"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "Create heatmap!<br>\n",
    "As viruses are extremelly alike the minimal value is set to 98.6"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "image/png": "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",
      "text/plain": [
       "<Figure size 1000x800 with 2 Axes>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "with open('ANI/fastani.out', 'r') as file:\n",
    "    content = file.readlines()\n",
    "\n",
    "# Extracting the necessary data from the file\n",
    "data = []\n",
    "for line in content:\n",
    "    parts = line.strip().split('\\t')\n",
    "    genome1 = parts[0].split('/')[-1]\n",
    "    genome2 = parts[1].split('/')[-1]\n",
    "    ani_value = float(parts[2])\n",
    "    data.append([genome1, genome2, ani_value])\n",
    "\n",
    "# Creating a DataFrame\n",
    "df = pd.DataFrame(data, columns=['Genome1', 'Genome2', 'ANI'])\n",
    "\n",
    "# Creating a pivot table for the heatmap\n",
    "pivot_table = df.pivot(index='Genome1', columns='Genome2', values='ANI')\n",
    "\n",
    "# Plotting the heatmap\n",
    "plt.figure(figsize=(10, 8))\n",
    "heatmap = sns.heatmap(pivot_table, annot=False, cmap='YlGnBu', cbar_kws={'label': 'ANI Percentage'}, vmin=98.6, vmax=100)\n",
    "#plt.xticks(rotation=45, ha='right')\n",
    "plt.xticks(rotation=90)\n",
    "plt.yticks(rotation=0)\n",
    "plt.xlabel(None)\n",
    "plt.ylabel(None)\n",
    "plt.savefig('images/fastANI_heatmap_x90.png', dpi=600, bbox_inches='tight')"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## **Part 6. Phylogenomics**"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "### **Part 6.1. Preparation**"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "Make directory for all the `phylogenomics` data"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 30,
   "metadata": {},
   "outputs": [],
   "source": [
    "! mkdir phylogenomics\n",
    "! mkdir phylogenomics/data/\n",
    "! mkdir phylogenomics/data/proteomes/"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "Copy densoviruses amino acid sequences produces by `prokka` to `phylogenomics/data/proteomes/` directory"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 31,
   "metadata": {},
   "outputs": [],
   "source": [
    "! cp prokka_results/*/*_annot.faa phylogenomics/data/proteomes/"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "Fetch the list of accession numbers of all the densoviruses from the `RefSeq` database<br>\n",
    "\n",
    "**NB!** Date of accession is Aug 2024"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 32,
   "metadata": {},
   "outputs": [],
   "source": [
    "! esearch -db nucleotide -query '(\"Ambidensovirus\"[Organism] OR \"densovirus\"[All Fields]) AND \"complete genome\"[All Fields] AND (viruses[filter] AND refseq[filter] AND (\"3000\"[SLEN] : \"6500\"[SLEN]))' | efetch -format acc > phylogenomics/data/accession_numbers.txt"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 33,
   "metadata": {},
   "outputs": [],
   "source": [
    "! echo \"MW628494.1\" >> phylogenomics/data/accession_numbers.txt"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "Import `Entrez` from `Biopython` and `os`"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 34,
   "metadata": {},
   "outputs": [],
   "source": [
    "from Bio import Entrez\n",
    "import os"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "Set the email (NCBI requires it for identification)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 35,
   "metadata": {},
   "outputs": [],
   "source": [
    "Entrez.email = \"iljapopov17@gmail.com\""
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "Function that will download amino acid sequences from the entries from the list of accession numbers"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 36,
   "metadata": {},
   "outputs": [],
   "source": [
    "# Function to download proteomes from NCBI\n",
    "def download_proteome(accession, output_dir):\n",
    "    # EFetch to download the fasta_cds_aa file\n",
    "    handle = Entrez.efetch(db=\"nuccore\", id=accession, rettype=\"fasta_cds_aa\", retmode=\"text\")\n",
    "    data = handle.read()\n",
    "    handle.close()\n",
    "    \n",
    "    # Save the fasta data\n",
    "    output_file = os.path.join(output_dir, f\"{accession}_proteome.fasta\")\n",
    "    with open(output_file, \"w\") as f:\n",
    "        f.write(data)\n",
    "    print(f\"Downloaded: {accession}\")\n",
    "\n",
    "# Function to process the list of accession numbers\n",
    "def process_accession_list(file_path, output_dir):\n",
    "    # Ensure the output directory exists\n",
    "    os.makedirs(output_dir, exist_ok=True)\n",
    "    \n",
    "    # Read the accession numbers from the file\n",
    "    with open(file_path, \"r\") as file:\n",
    "        accessions = [line.strip() for line in file.readlines()]\n",
    "    \n",
    "    # Download each proteome\n",
    "    for accession in accessions:\n",
    "        download_proteome(accession, output_dir)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "Run the function to download amino acid sequences of densoviruses from `RefSeq` database"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [],
   "source": [
    "# Input: File with accession numbers and output directory\n",
    "accession_file = \"phylogenomics/data/accession_numbers.txt\"\n",
    "output_directory = \"phylogenomics/data/proteomes/\"\n",
    "\n",
    "# Run the process\n",
    "process_accession_list(accession_file, output_directory)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "### **Part 6.2. \"Sequence brushing\"**"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "Let's look at the sequences' names"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 38,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      ">KCBIPIHA_00001 structural protein [Blattella germanica densovirus-like virus]\n"
     ]
    }
   ],
   "source": [
    "! head -1 phylogenomics/data/proteomes/Densovirus_D1_annot.faa"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 39,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      ">lcl|NC_000936.1_prot_NP_051016.1_1 [locus_tag=PfdVgp1] [db_xref=GeneID:1457168] [protein=structural protein] [protein_id=NP_051016.1] [location=202..930] [gbkey=CDS]\n"
     ]
    }
   ],
   "source": [
    "! head -1 phylogenomics/data/proteomes/NC_000936.1_proteome.fasta"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "It looks informative, but awful. We will use `Proteinortho` later. And this tool will not take these sequences as the input because of their seqs' names. We must rename them. It will be not a simple process. That's why it will be in several steps."
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "Make directory to store renamed sequences<br>\n",
    "And make enough directories to allow step-by-step process"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 40,
   "metadata": {},
   "outputs": [],
   "source": [
    "! mkdir phylogenomics/data/proteomes_renamed/\n",
    "! mkdir phylogenomics/data/proteomes_renamed/round_1/\n",
    "! mkdir phylogenomics/data/proteomes_renamed/round_2/\n",
    "! mkdir phylogenomics/data/proteomes_renamed/round_3/"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "#### **Part 6.2.1. Round 1**"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "For the 1st round we will:\n",
    "\n",
    "- Make this `>KCBIPIHA_00001 structural protein [Blattella germanica densovirus-like virus]` look like this `>structural protein`<br>\n",
    "- Make this `>lcl|NC_000936.1_prot_NP_051016.1_1 [locus_tag=PfdVgp1] [db_xref=GeneID:1457168] [protein=structural protein] [protein_id=NP_051016.1] [location=202..930] [gbkey=CDS]` look like this `>structural protein`\n",
    "\n",
    "So we will delete everything except the gene names"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 41,
   "metadata": {},
   "outputs": [],
   "source": [
    "%%bash\n",
    "\n",
    "for file in phylogenomics/data/proteomes/*_annot.faa\n",
    "do sed -E 's/^>([^ ]+) ([^[]+)\\[.*\\]/>\\2/' $file > phylogenomics/data/proteomes_renamed/round_1/$(basename \"$file\" _annot.faa)_renamed.fa\n",
    "done"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 42,
   "metadata": {},
   "outputs": [],
   "source": [
    "%%bash\n",
    "\n",
    "for file in phylogenomics/data/proteomes/*.fasta\n",
    "do sed -E 's/^>.*\\[protein=([^]]+)\\].*/>\\1/' $file > phylogenomics/data/proteomes_renamed/round_1/$(basename \"$file\" _proteome.fasta)_renamed.fa\n",
    "done"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "Let's check if `sed` did good"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 43,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      ">structural protein \n"
     ]
    }
   ],
   "source": [
    "! head -1 phylogenomics/data/proteomes_renamed/round_1/Densovirus_D1_renamed.fa"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 44,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      ">structural protein\n"
     ]
    }
   ],
   "source": [
    "! head -1 phylogenomics/data/proteomes_renamed/round_1/NC_000936.1_renamed.fa"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "Perfect. Let's move on the 2nd round"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "#### **Part 6.2.2. Round 2**"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "First we need to install `BBMap`"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "Download `.tar.gz` archive"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [],
   "source": [
    "! wget https://sourceforge.net/projects/bbmap/files/latest/download -O BBMap.tar.gz"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "Unpack it"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [],
   "source": [
    "! tar -xvzf BBMap.tar.gz"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "Delete the initial archive"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [],
   "source": [
    "! rm -rf BBMap.tar.gz"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "For the 2nd round we will:\n",
    "\n",
    "- Make this `>structural protein` look like this: `>NC_000936.1_structural protein`<br>\n",
    "\n",
    "So, we will make each sequence unique by binding the information of which exactly virus this gene is"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [],
   "source": [
    "%%bash\n",
    "\n",
    "for file in phylogenomics/data/proteomes_renamed/round_1/*_renamed.fa\n",
    "do export species_name=$(basename $file _renamed.fa)\n",
    "bbmap/rename.sh in=$file prefix=\"$species_name\" addprefix=true out=phylogenomics/data/proteomes_renamed/round_2/$species_name.fa ignorejunk=true ;\n",
    "done"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "Let's check if `BBMap` did good"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 47,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      ">NC_000936.1_structural protein\n"
     ]
    }
   ],
   "source": [
    "! head -1 phylogenomics/data/proteomes_renamed/round_2/NC_000936.1.fa"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "Perfect"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "#### **Part 6.2.3. Round 3**"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "For the 3rd round we will:\n",
    "\n",
    "- Make this `>Densovirus_P3_structural protein` look like this: `>Densovirus_P3_structural_protein_`<br>\n",
    "\n",
    "So, we will replace each space with underline"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 48,
   "metadata": {},
   "outputs": [],
   "source": [
    "%%bash\n",
    "\n",
    "for file in phylogenomics/data/proteomes_renamed/round_2/*.fa\n",
    "do sed -E '/^>/ s/ /_/g' $file > phylogenomics/data/proteomes_renamed/round_3/$(basename \"$file\")\n",
    "done"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "Let's check if `sed` did good"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 49,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      ">Densovirus_P3_structural_protein_\n"
     ]
    }
   ],
   "source": [
    "! head -1 phylogenomics/data/proteomes_renamed/round_3/Densovirus_P3.fa"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "Perfect"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "#### **Part 6.2.4. Round 4**"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "Import `SeqIO` from `Biopython`"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 50,
   "metadata": {},
   "outputs": [],
   "source": [
    "from Bio import SeqIO"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "For the 4th round we will:\n",
    "\n",
    "- Make this `>Densovirus_P3_structural protein` look like this: `>Densovirus_P3_structural_protein__1`<br>\n",
    "\n",
    "So, we will enumerate genes. Some viruses has several structural or nonstructural proteins without enumeration. It will be bad further, so we must enumerate them."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 51,
   "metadata": {},
   "outputs": [],
   "source": [
    "def rename_sequences(input_dir, output_dir):\n",
    "    # Create the output directory if it doesn't exist\n",
    "    os.makedirs(output_dir, exist_ok=True)\n",
    "    \n",
    "    # Loop through each FASTA file in the directory\n",
    "    for filename in os.listdir(input_dir):\n",
    "        if filename.endswith(\".fasta\") or filename.endswith(\".fa\"):\n",
    "            filepath = os.path.join(input_dir, filename)\n",
    "            output_filepath = os.path.join(output_dir, filename)\n",
    "\n",
    "            # Dictionary to keep track of sequence names and their counts\n",
    "            name_count = {}\n",
    "            \n",
    "            # List to store the updated sequences\n",
    "            updated_sequences = []\n",
    "            \n",
    "            # Read the FASTA file\n",
    "            for record in SeqIO.parse(filepath, \"fasta\"):\n",
    "                name = record.id\n",
    "                \n",
    "                # If the name is already seen, append a count suffix\n",
    "                if name in name_count:\n",
    "                    name_count[name] += 1\n",
    "                    new_name = f\"{name}_{name_count[name]}\"\n",
    "                else:\n",
    "                    name_count[name] = 1\n",
    "                    new_name = f\"{name}_1\"\n",
    "                \n",
    "                # Update the record ID\n",
    "                record.id = new_name\n",
    "                record.description = \"\"  # Optionally clear the description\n",
    "                \n",
    "                # Store the updated record\n",
    "                updated_sequences.append(record)\n",
    "\n",
    "            # Write the updated sequences to a new FASTA file\n",
    "            with open(output_filepath, \"w\") as output_handle:\n",
    "                SeqIO.write(updated_sequences, output_handle, \"fasta\")"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "Run the function above"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 52,
   "metadata": {},
   "outputs": [],
   "source": [
    "input_dir = \"phylogenomics/data/proteomes_renamed/round_3\"\n",
    "output_dir = \"phylogenomics/data/proteomes_renamed/round_4\"\n",
    "rename_sequences(input_dir, output_dir)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "Let's check if function `rename_sequences()` did good"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 53,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      ">Densovirus_P3_structural_protein__1\n"
     ]
    }
   ],
   "source": [
    "! head -1 phylogenomics/data/proteomes_renamed/round_4/Densovirus_P3.fa"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "Perfect"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "**Disclaimer**<br>\n",
    "Unfortunatelly, we have to delete the `NC_004284.1 Junonia coenia densovirus, complete genome` from the analysis, because `Proteinortho` complains that this exactly sequence has some inapropriate letters in it."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 54,
   "metadata": {},
   "outputs": [],
   "source": [
    "! rm -rf phylogenomics/data/proteomes_renamed/round_4/NC_004284.1.fa"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "### **Part 6.3. Launching `Proteinortho` to find the orhologs**"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "Launch `Proteinortho`"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [],
   "source": [
    "! proteinortho phylogenomics/data/proteomes_renamed/round_4/*.fa -cpus=24"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "Unfortunatelly, `Proteinortho` does not let you to redirect output to any directory you want...<br>\n",
    "That's why we move the files ourselves!<br>\n",
    "For that first we create a directory - `phylogenomics/proteinortho_output/`"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 56,
   "metadata": {},
   "outputs": [],
   "source": [
    "! mkdir phylogenomics/proteinortho_output/"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "Then we just move any `myproject*` file to `phylogenomics/proteinortho_output/` directory"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 57,
   "metadata": {},
   "outputs": [],
   "source": [
    "! mv myproject* phylogenomics/proteinortho_output/"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "At this point it is important to take time for manual exploration of the `phylogenomics/proteinortho_output/myproject.proteinortho.tsv` file. To find any single copy orthologs or to find any other interesting genes."
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "### **Part 6.4. Extracting orthologs from `Proteinortho` output**"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "Import `pandas`"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 58,
   "metadata": {},
   "outputs": [],
   "source": [
    "import pandas as pd"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "Read the `phylogenomics/proteinortho_output/myproject.proteinortho.tsv` file into a DataFrame"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 59,
   "metadata": {},
   "outputs": [],
   "source": [
    "df_all = pd.read_csv('phylogenomics/proteinortho_output/myproject.proteinortho.tsv', sep='\\t')"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "Drop all the rows except 0, 1 and 3"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 60,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/html": [
       "<div>\n",
       "<style scoped>\n",
       "    .dataframe tbody tr th:only-of-type {\n",
       "        vertical-align: middle;\n",
       "    }\n",
       "\n",
       "    .dataframe tbody tr th {\n",
       "        vertical-align: top;\n",
       "    }\n",
       "\n",
       "    .dataframe thead th {\n",
       "        text-align: right;\n",
       "    }\n",
       "</style>\n",
       "<table border=\"1\" class=\"dataframe\">\n",
       "  <thead>\n",
       "    <tr style=\"text-align: right;\">\n",
       "      <th></th>\n",
       "      <th># Species</th>\n",
       "      <th>Genes</th>\n",
       "      <th>Alg.-Conn.</th>\n",
       "      <th>Densovirus_D1.fa</th>\n",
       "      <th>Densovirus_D2.fa</th>\n",
       "      <th>Densovirus_D3.fa</th>\n",
       "      <th>Densovirus_D4.fa</th>\n",
       "      <th>Densovirus_D5.fa</th>\n",
       "      <th>Densovirus_P1.fa</th>\n",
       "      <th>Densovirus_P2.fa</th>\n",
       "      <th>...</th>\n",
       "      <th>NC_077038.1.fa</th>\n",
       "      <th>NC_077050.1.fa</th>\n",
       "      <th>NC_077051.1.fa</th>\n",
       "      <th>NC_077065.1.fa</th>\n",
       "      <th>NC_077069.1.fa</th>\n",
       "      <th>NC_077079.1.fa</th>\n",
       "      <th>NC_077080.1.fa</th>\n",
       "      <th>NC_077081.1.fa</th>\n",
       "      <th>NC_077082.1.fa</th>\n",
       "      <th>NC_077083.1.fa</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>0</th>\n",
       "      <td>45</td>\n",
       "      <td>45</td>\n",
       "      <td>0.0222</td>\n",
       "      <td>Densovirus_D1_non-structural_protein__1</td>\n",
       "      <td>Densovirus_D2_non-structural_protein__1</td>\n",
       "      <td>Densovirus_D3_non-structural_protein__1</td>\n",
       "      <td>Densovirus_D4_non-structural_protein__1</td>\n",
       "      <td>Densovirus_D5_non-structural_protein__1</td>\n",
       "      <td>Densovirus_P1_non-structural_protein__1</td>\n",
       "      <td>Densovirus_P2_non-structural_protein__1</td>\n",
       "      <td>...</td>\n",
       "      <td>NC_077038.1_nonstructural_protein_2</td>\n",
       "      <td>NC_077050.1_putative_nonstructural_protein_1</td>\n",
       "      <td>*</td>\n",
       "      <td>NC_077065.1_hypothetical_protein_3</td>\n",
       "      <td>NC_077069.1_nonstructural_protein_1_1</td>\n",
       "      <td>NC_077079.1_nonstructural_protein_1</td>\n",
       "      <td>NC_077080.1_nonstructural_protein_1</td>\n",
       "      <td>NC_077081.1_nonstructural_protein_1</td>\n",
       "      <td>NC_077082.1_nonstructural_protein_1</td>\n",
       "      <td>NC_077083.1_nonstructural_protein_1</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1</th>\n",
       "      <td>28</td>\n",
       "      <td>43</td>\n",
       "      <td>0.1920</td>\n",
       "      <td>Densovirus_D1_ORF1__1</td>\n",
       "      <td>Densovirus_D2_ORF1__1</td>\n",
       "      <td>Densovirus_D3_ORF1__1</td>\n",
       "      <td>Densovirus_D4_ORF1__1</td>\n",
       "      <td>Densovirus_D5_ORF1__1</td>\n",
       "      <td>Densovirus_P1_ORF1__1</td>\n",
       "      <td>Densovirus_P2_ORF1__1</td>\n",
       "      <td>...</td>\n",
       "      <td>*</td>\n",
       "      <td>*</td>\n",
       "      <td>*</td>\n",
       "      <td>NC_077065.1_hypothetical_protein_2</td>\n",
       "      <td>*</td>\n",
       "      <td>NC_077079.1_VP1_1</td>\n",
       "      <td>NC_077080.1_VP1_1</td>\n",
       "      <td>*</td>\n",
       "      <td>NC_077082.1_VP1_1</td>\n",
       "      <td>*</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2</th>\n",
       "      <td>12</td>\n",
       "      <td>12</td>\n",
       "      <td>1.0000</td>\n",
       "      <td>Densovirus_D1_structural_protein__1</td>\n",
       "      <td>Densovirus_D2_structural_protein__1</td>\n",
       "      <td>Densovirus_D3_structural_protein__1</td>\n",
       "      <td>Densovirus_D4_structural_protein__1</td>\n",
       "      <td>Densovirus_D5_structural_protein__1</td>\n",
       "      <td>Densovirus_P1_structural_protein__1</td>\n",
       "      <td>Densovirus_P2_structural_protein__1</td>\n",
       "      <td>...</td>\n",
       "      <td>*</td>\n",
       "      <td>*</td>\n",
       "      <td>*</td>\n",
       "      <td>*</td>\n",
       "      <td>*</td>\n",
       "      <td>*</td>\n",
       "      <td>*</td>\n",
       "      <td>*</td>\n",
       "      <td>*</td>\n",
       "      <td>*</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "<p>3 rows × 57 columns</p>\n",
       "</div>"
      ],
      "text/plain": [
       "   # Species  Genes  Alg.-Conn.                         Densovirus_D1.fa  \\\n",
       "0         45     45      0.0222  Densovirus_D1_non-structural_protein__1   \n",
       "1         28     43      0.1920                    Densovirus_D1_ORF1__1   \n",
       "2         12     12      1.0000      Densovirus_D1_structural_protein__1   \n",
       "\n",
       "                          Densovirus_D2.fa  \\\n",
       "0  Densovirus_D2_non-structural_protein__1   \n",
       "1                    Densovirus_D2_ORF1__1   \n",
       "2      Densovirus_D2_structural_protein__1   \n",
       "\n",
       "                          Densovirus_D3.fa  \\\n",
       "0  Densovirus_D3_non-structural_protein__1   \n",
       "1                    Densovirus_D3_ORF1__1   \n",
       "2      Densovirus_D3_structural_protein__1   \n",
       "\n",
       "                          Densovirus_D4.fa  \\\n",
       "0  Densovirus_D4_non-structural_protein__1   \n",
       "1                    Densovirus_D4_ORF1__1   \n",
       "2      Densovirus_D4_structural_protein__1   \n",
       "\n",
       "                          Densovirus_D5.fa  \\\n",
       "0  Densovirus_D5_non-structural_protein__1   \n",
       "1                    Densovirus_D5_ORF1__1   \n",
       "2      Densovirus_D5_structural_protein__1   \n",
       "\n",
       "                          Densovirus_P1.fa  \\\n",
       "0  Densovirus_P1_non-structural_protein__1   \n",
       "1                    Densovirus_P1_ORF1__1   \n",
       "2      Densovirus_P1_structural_protein__1   \n",
       "\n",
       "                          Densovirus_P2.fa  ...  \\\n",
       "0  Densovirus_P2_non-structural_protein__1  ...   \n",
       "1                    Densovirus_P2_ORF1__1  ...   \n",
       "2      Densovirus_P2_structural_protein__1  ...   \n",
       "\n",
       "                        NC_077038.1.fa  \\\n",
       "0  NC_077038.1_nonstructural_protein_2   \n",
       "1                                    *   \n",
       "2                                    *   \n",
       "\n",
       "                                 NC_077050.1.fa NC_077051.1.fa  \\\n",
       "0  NC_077050.1_putative_nonstructural_protein_1              *   \n",
       "1                                             *              *   \n",
       "2                                             *              *   \n",
       "\n",
       "                       NC_077065.1.fa                         NC_077069.1.fa  \\\n",
       "0  NC_077065.1_hypothetical_protein_3  NC_077069.1_nonstructural_protein_1_1   \n",
       "1  NC_077065.1_hypothetical_protein_2                                      *   \n",
       "2                                   *                                      *   \n",
       "\n",
       "                        NC_077079.1.fa                       NC_077080.1.fa  \\\n",
       "0  NC_077079.1_nonstructural_protein_1  NC_077080.1_nonstructural_protein_1   \n",
       "1                    NC_077079.1_VP1_1                    NC_077080.1_VP1_1   \n",
       "2                                    *                                    *   \n",
       "\n",
       "                        NC_077081.1.fa                       NC_077082.1.fa  \\\n",
       "0  NC_077081.1_nonstructural_protein_1  NC_077082.1_nonstructural_protein_1   \n",
       "1                                    *                    NC_077082.1_VP1_1   \n",
       "2                                    *                                    *   \n",
       "\n",
       "                        NC_077083.1.fa  \n",
       "0  NC_077083.1_nonstructural_protein_1  \n",
       "1                                    *  \n",
       "2                                    *  \n",
       "\n",
       "[3 rows x 57 columns]"
      ]
     },
     "execution_count": 60,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "for i in range(3, 23 + 1):\n",
    "    df_all = df_all.drop(index=[i])\n",
    "\n",
    "df_all"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "These are the orthologs we will work with:\n",
    "\n",
    "- `Non-structural protein` is presented in 44 viruses (10 from this study and 34 from `RefSeq`)\n",
    "- `ORF1` or `structural protein` is presented in 27 viruses (10 from this study and 17 from `RefSeq`), but in 5 viruses it is not single-copy. We will delete them from our dataset. Which makes this gene to be presented only in 22 viruses (10 from this study and 12 from `RefSeq`)\n",
    "- Another `structural protein` is presented in 11 viruses (10 from this study and 1 from `RefSeq`). It is interesting because its `Alg.-Conn.` parameter = `1.0000`. We must investigate it."
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "Delete all non-SCOs"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 61,
   "metadata": {},
   "outputs": [],
   "source": [
    "df_all.loc[1] = df_all.loc[1].apply(lambda x: x.replace(',', '*') if isinstance(x, str) else x)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "#### **Part 6.4.1. Non-structural protein**"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 62,
   "metadata": {},
   "outputs": [],
   "source": [
    "df_nsp = df_all.drop(index=[1,2])"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 63,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/html": [
       "<div>\n",
       "<style scoped>\n",
       "    .dataframe tbody tr th:only-of-type {\n",
       "        vertical-align: middle;\n",
       "    }\n",
       "\n",
       "    .dataframe tbody tr th {\n",
       "        vertical-align: top;\n",
       "    }\n",
       "\n",
       "    .dataframe thead th {\n",
       "        text-align: right;\n",
       "    }\n",
       "</style>\n",
       "<table border=\"1\" class=\"dataframe\">\n",
       "  <thead>\n",
       "    <tr style=\"text-align: right;\">\n",
       "      <th></th>\n",
       "      <th># Species</th>\n",
       "      <th>Genes</th>\n",
       "      <th>Alg.-Conn.</th>\n",
       "      <th>Densovirus_D1.fa</th>\n",
       "      <th>Densovirus_D2.fa</th>\n",
       "      <th>Densovirus_D3.fa</th>\n",
       "      <th>Densovirus_D4.fa</th>\n",
       "      <th>Densovirus_D5.fa</th>\n",
       "      <th>Densovirus_P1.fa</th>\n",
       "      <th>Densovirus_P2.fa</th>\n",
       "      <th>...</th>\n",
       "      <th>NC_077038.1.fa</th>\n",
       "      <th>NC_077050.1.fa</th>\n",
       "      <th>NC_077051.1.fa</th>\n",
       "      <th>NC_077065.1.fa</th>\n",
       "      <th>NC_077069.1.fa</th>\n",
       "      <th>NC_077079.1.fa</th>\n",
       "      <th>NC_077080.1.fa</th>\n",
       "      <th>NC_077081.1.fa</th>\n",
       "      <th>NC_077082.1.fa</th>\n",
       "      <th>NC_077083.1.fa</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>0</th>\n",
       "      <td>45</td>\n",
       "      <td>45</td>\n",
       "      <td>0.0222</td>\n",
       "      <td>Densovirus_D1_non-structural_protein__1</td>\n",
       "      <td>Densovirus_D2_non-structural_protein__1</td>\n",
       "      <td>Densovirus_D3_non-structural_protein__1</td>\n",
       "      <td>Densovirus_D4_non-structural_protein__1</td>\n",
       "      <td>Densovirus_D5_non-structural_protein__1</td>\n",
       "      <td>Densovirus_P1_non-structural_protein__1</td>\n",
       "      <td>Densovirus_P2_non-structural_protein__1</td>\n",
       "      <td>...</td>\n",
       "      <td>NC_077038.1_nonstructural_protein_2</td>\n",
       "      <td>NC_077050.1_putative_nonstructural_protein_1</td>\n",
       "      <td>*</td>\n",
       "      <td>NC_077065.1_hypothetical_protein_3</td>\n",
       "      <td>NC_077069.1_nonstructural_protein_1_1</td>\n",
       "      <td>NC_077079.1_nonstructural_protein_1</td>\n",
       "      <td>NC_077080.1_nonstructural_protein_1</td>\n",
       "      <td>NC_077081.1_nonstructural_protein_1</td>\n",
       "      <td>NC_077082.1_nonstructural_protein_1</td>\n",
       "      <td>NC_077083.1_nonstructural_protein_1</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "<p>1 rows × 57 columns</p>\n",
       "</div>"
      ],
      "text/plain": [
       "   # Species  Genes  Alg.-Conn.                         Densovirus_D1.fa  \\\n",
       "0         45     45      0.0222  Densovirus_D1_non-structural_protein__1   \n",
       "\n",
       "                          Densovirus_D2.fa  \\\n",
       "0  Densovirus_D2_non-structural_protein__1   \n",
       "\n",
       "                          Densovirus_D3.fa  \\\n",
       "0  Densovirus_D3_non-structural_protein__1   \n",
       "\n",
       "                          Densovirus_D4.fa  \\\n",
       "0  Densovirus_D4_non-structural_protein__1   \n",
       "\n",
       "                          Densovirus_D5.fa  \\\n",
       "0  Densovirus_D5_non-structural_protein__1   \n",
       "\n",
       "                          Densovirus_P1.fa  \\\n",
       "0  Densovirus_P1_non-structural_protein__1   \n",
       "\n",
       "                          Densovirus_P2.fa  ...  \\\n",
       "0  Densovirus_P2_non-structural_protein__1  ...   \n",
       "\n",
       "                        NC_077038.1.fa  \\\n",
       "0  NC_077038.1_nonstructural_protein_2   \n",
       "\n",
       "                                 NC_077050.1.fa NC_077051.1.fa  \\\n",
       "0  NC_077050.1_putative_nonstructural_protein_1              *   \n",
       "\n",
       "                       NC_077065.1.fa                         NC_077069.1.fa  \\\n",
       "0  NC_077065.1_hypothetical_protein_3  NC_077069.1_nonstructural_protein_1_1   \n",
       "\n",
       "                        NC_077079.1.fa                       NC_077080.1.fa  \\\n",
       "0  NC_077079.1_nonstructural_protein_1  NC_077080.1_nonstructural_protein_1   \n",
       "\n",
       "                        NC_077081.1.fa                       NC_077082.1.fa  \\\n",
       "0  NC_077081.1_nonstructural_protein_1  NC_077082.1_nonstructural_protein_1   \n",
       "\n",
       "                        NC_077083.1.fa  \n",
       "0  NC_077083.1_nonstructural_protein_1  \n",
       "\n",
       "[1 rows x 57 columns]"
      ]
     },
     "execution_count": 63,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "df_nsp"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "Delete every entry with `*`"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 64,
   "metadata": {},
   "outputs": [],
   "source": [
    "df_nsp_cols_to_drop = df_nsp.columns[df_nsp.apply(lambda col: col.astype(str).str.contains(r'\\*', regex=True).any())]"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "Examine which entries will be dropped"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 65,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "Index(['NC_011317.1.fa', 'NC_011545.2.fa', 'NC_012636.1.fa', 'NC_014357.1.fa',\n",
       "       'NC_015115.1.fa', 'NC_020927.1.fa', 'NC_038376.1.fa', 'NC_077018.1.fa',\n",
       "       'NC_077051.1.fa'],\n",
       "      dtype='object')"
      ]
     },
     "execution_count": 65,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "df_nsp_cols_to_drop"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "Save a separate `.tsv` file"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 66,
   "metadata": {},
   "outputs": [],
   "source": [
    "df_nsp_cleaned = df_nsp.drop(columns=df_nsp_cols_to_drop)\n",
    "df_nsp_cleaned.to_csv('phylogenomics/proteinortho_output/NSPs.tsv', sep='\\t', index=False)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "#### **Part 6.4.2. Structural protein 1**"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 67,
   "metadata": {},
   "outputs": [],
   "source": [
    "df_sp = df_all.drop(index=[0, 2])"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 68,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/html": [
       "<div>\n",
       "<style scoped>\n",
       "    .dataframe tbody tr th:only-of-type {\n",
       "        vertical-align: middle;\n",
       "    }\n",
       "\n",
       "    .dataframe tbody tr th {\n",
       "        vertical-align: top;\n",
       "    }\n",
       "\n",
       "    .dataframe thead th {\n",
       "        text-align: right;\n",
       "    }\n",
       "</style>\n",
       "<table border=\"1\" class=\"dataframe\">\n",
       "  <thead>\n",
       "    <tr style=\"text-align: right;\">\n",
       "      <th></th>\n",
       "      <th># Species</th>\n",
       "      <th>Genes</th>\n",
       "      <th>Alg.-Conn.</th>\n",
       "      <th>Densovirus_D1.fa</th>\n",
       "      <th>Densovirus_D2.fa</th>\n",
       "      <th>Densovirus_D3.fa</th>\n",
       "      <th>Densovirus_D4.fa</th>\n",
       "      <th>Densovirus_D5.fa</th>\n",
       "      <th>Densovirus_P1.fa</th>\n",
       "      <th>Densovirus_P2.fa</th>\n",
       "      <th>...</th>\n",
       "      <th>NC_077038.1.fa</th>\n",
       "      <th>NC_077050.1.fa</th>\n",
       "      <th>NC_077051.1.fa</th>\n",
       "      <th>NC_077065.1.fa</th>\n",
       "      <th>NC_077069.1.fa</th>\n",
       "      <th>NC_077079.1.fa</th>\n",
       "      <th>NC_077080.1.fa</th>\n",
       "      <th>NC_077081.1.fa</th>\n",
       "      <th>NC_077082.1.fa</th>\n",
       "      <th>NC_077083.1.fa</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>1</th>\n",
       "      <td>28</td>\n",
       "      <td>43</td>\n",
       "      <td>0.192</td>\n",
       "      <td>Densovirus_D1_ORF1__1</td>\n",
       "      <td>Densovirus_D2_ORF1__1</td>\n",
       "      <td>Densovirus_D3_ORF1__1</td>\n",
       "      <td>Densovirus_D4_ORF1__1</td>\n",
       "      <td>Densovirus_D5_ORF1__1</td>\n",
       "      <td>Densovirus_P1_ORF1__1</td>\n",
       "      <td>Densovirus_P2_ORF1__1</td>\n",
       "      <td>...</td>\n",
       "      <td>*</td>\n",
       "      <td>*</td>\n",
       "      <td>*</td>\n",
       "      <td>NC_077065.1_hypothetical_protein_2</td>\n",
       "      <td>*</td>\n",
       "      <td>NC_077079.1_VP1_1</td>\n",
       "      <td>NC_077080.1_VP1_1</td>\n",
       "      <td>*</td>\n",
       "      <td>NC_077082.1_VP1_1</td>\n",
       "      <td>*</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "<p>1 rows × 57 columns</p>\n",
       "</div>"
      ],
      "text/plain": [
       "   # Species  Genes  Alg.-Conn.       Densovirus_D1.fa       Densovirus_D2.fa  \\\n",
       "1         28     43       0.192  Densovirus_D1_ORF1__1  Densovirus_D2_ORF1__1   \n",
       "\n",
       "        Densovirus_D3.fa       Densovirus_D4.fa       Densovirus_D5.fa  \\\n",
       "1  Densovirus_D3_ORF1__1  Densovirus_D4_ORF1__1  Densovirus_D5_ORF1__1   \n",
       "\n",
       "        Densovirus_P1.fa       Densovirus_P2.fa  ... NC_077038.1.fa  \\\n",
       "1  Densovirus_P1_ORF1__1  Densovirus_P2_ORF1__1  ...              *   \n",
       "\n",
       "  NC_077050.1.fa NC_077051.1.fa                      NC_077065.1.fa  \\\n",
       "1              *              *  NC_077065.1_hypothetical_protein_2   \n",
       "\n",
       "  NC_077069.1.fa     NC_077079.1.fa     NC_077080.1.fa NC_077081.1.fa  \\\n",
       "1              *  NC_077079.1_VP1_1  NC_077080.1_VP1_1              *   \n",
       "\n",
       "      NC_077082.1.fa NC_077083.1.fa  \n",
       "1  NC_077082.1_VP1_1              *  \n",
       "\n",
       "[1 rows x 57 columns]"
      ]
     },
     "execution_count": 68,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "df_sp"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "Delete every entry with `*`"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 69,
   "metadata": {},
   "outputs": [],
   "source": [
    "df_sp_cols_to_drop = df_sp.columns[df_sp.apply(lambda col: col.astype(str).str.contains(r'\\*', regex=True).any())]"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "Examine which entries will be dropped"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 70,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "Index(['NC_001899.1.fa', 'NC_003346.1.fa', 'NC_004286.1.fa', 'NC_004287.1.fa',\n",
       "       'NC_004288.1.fa', 'NC_005041.2.fa', 'NC_005341.1.fa', 'NC_006555.1.fa',\n",
       "       'NC_011317.1.fa', 'NC_011545.2.fa', 'NC_012636.1.fa', 'NC_014357.1.fa',\n",
       "       'NC_015115.1.fa', 'NC_015718.1.fa', 'NC_018399.1.fa', 'NC_018450.1.fa',\n",
       "       'NC_019492.1.fa', 'NC_020927.1.fa', 'NC_022564.1.fa', 'NC_026943.1.fa',\n",
       "       'NC_030296.1.fa', 'NC_038376.1.fa', 'NC_040626.1.fa', 'NC_075118.1.fa',\n",
       "       'NC_077018.1.fa', 'NC_077036.1.fa', 'NC_077038.1.fa', 'NC_077050.1.fa',\n",
       "       'NC_077051.1.fa', 'NC_077069.1.fa', 'NC_077081.1.fa', 'NC_077083.1.fa'],\n",
       "      dtype='object')"
      ]
     },
     "execution_count": 70,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "df_sp_cols_to_drop"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "Save a separate `.tsv` file"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 71,
   "metadata": {},
   "outputs": [],
   "source": [
    "df_sp_cleaned = df_sp.drop(columns=df_sp_cols_to_drop)\n",
    "df_sp_cleaned.to_csv('phylogenomics/proteinortho_output/SPs.tsv', sep='\\t', index=False)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "#### **Part 6.4.3. Non-structural protein + structural protein 1**"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "We also need a `.tsv` file with NSP and SP"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "Delete every entry with `*`"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 72,
   "metadata": {},
   "outputs": [],
   "source": [
    "df_nsp_sp = df_all.drop(index=[2])"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 73,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/html": [
       "<div>\n",
       "<style scoped>\n",
       "    .dataframe tbody tr th:only-of-type {\n",
       "        vertical-align: middle;\n",
       "    }\n",
       "\n",
       "    .dataframe tbody tr th {\n",
       "        vertical-align: top;\n",
       "    }\n",
       "\n",
       "    .dataframe thead th {\n",
       "        text-align: right;\n",
       "    }\n",
       "</style>\n",
       "<table border=\"1\" class=\"dataframe\">\n",
       "  <thead>\n",
       "    <tr style=\"text-align: right;\">\n",
       "      <th></th>\n",
       "      <th># Species</th>\n",
       "      <th>Genes</th>\n",
       "      <th>Alg.-Conn.</th>\n",
       "      <th>Densovirus_D1.fa</th>\n",
       "      <th>Densovirus_D2.fa</th>\n",
       "      <th>Densovirus_D3.fa</th>\n",
       "      <th>Densovirus_D4.fa</th>\n",
       "      <th>Densovirus_D5.fa</th>\n",
       "      <th>Densovirus_P1.fa</th>\n",
       "      <th>Densovirus_P2.fa</th>\n",
       "      <th>...</th>\n",
       "      <th>NC_077038.1.fa</th>\n",
       "      <th>NC_077050.1.fa</th>\n",
       "      <th>NC_077051.1.fa</th>\n",
       "      <th>NC_077065.1.fa</th>\n",
       "      <th>NC_077069.1.fa</th>\n",
       "      <th>NC_077079.1.fa</th>\n",
       "      <th>NC_077080.1.fa</th>\n",
       "      <th>NC_077081.1.fa</th>\n",
       "      <th>NC_077082.1.fa</th>\n",
       "      <th>NC_077083.1.fa</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>0</th>\n",
       "      <td>45</td>\n",
       "      <td>45</td>\n",
       "      <td>0.0222</td>\n",
       "      <td>Densovirus_D1_non-structural_protein__1</td>\n",
       "      <td>Densovirus_D2_non-structural_protein__1</td>\n",
       "      <td>Densovirus_D3_non-structural_protein__1</td>\n",
       "      <td>Densovirus_D4_non-structural_protein__1</td>\n",
       "      <td>Densovirus_D5_non-structural_protein__1</td>\n",
       "      <td>Densovirus_P1_non-structural_protein__1</td>\n",
       "      <td>Densovirus_P2_non-structural_protein__1</td>\n",
       "      <td>...</td>\n",
       "      <td>NC_077038.1_nonstructural_protein_2</td>\n",
       "      <td>NC_077050.1_putative_nonstructural_protein_1</td>\n",
       "      <td>*</td>\n",
       "      <td>NC_077065.1_hypothetical_protein_3</td>\n",
       "      <td>NC_077069.1_nonstructural_protein_1_1</td>\n",
       "      <td>NC_077079.1_nonstructural_protein_1</td>\n",
       "      <td>NC_077080.1_nonstructural_protein_1</td>\n",
       "      <td>NC_077081.1_nonstructural_protein_1</td>\n",
       "      <td>NC_077082.1_nonstructural_protein_1</td>\n",
       "      <td>NC_077083.1_nonstructural_protein_1</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1</th>\n",
       "      <td>28</td>\n",
       "      <td>43</td>\n",
       "      <td>0.1920</td>\n",
       "      <td>Densovirus_D1_ORF1__1</td>\n",
       "      <td>Densovirus_D2_ORF1__1</td>\n",
       "      <td>Densovirus_D3_ORF1__1</td>\n",
       "      <td>Densovirus_D4_ORF1__1</td>\n",
       "      <td>Densovirus_D5_ORF1__1</td>\n",
       "      <td>Densovirus_P1_ORF1__1</td>\n",
       "      <td>Densovirus_P2_ORF1__1</td>\n",
       "      <td>...</td>\n",
       "      <td>*</td>\n",
       "      <td>*</td>\n",
       "      <td>*</td>\n",
       "      <td>NC_077065.1_hypothetical_protein_2</td>\n",
       "      <td>*</td>\n",
       "      <td>NC_077079.1_VP1_1</td>\n",
       "      <td>NC_077080.1_VP1_1</td>\n",
       "      <td>*</td>\n",
       "      <td>NC_077082.1_VP1_1</td>\n",
       "      <td>*</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "<p>2 rows × 57 columns</p>\n",
       "</div>"
      ],
      "text/plain": [
       "   # Species  Genes  Alg.-Conn.                         Densovirus_D1.fa  \\\n",
       "0         45     45      0.0222  Densovirus_D1_non-structural_protein__1   \n",
       "1         28     43      0.1920                    Densovirus_D1_ORF1__1   \n",
       "\n",
       "                          Densovirus_D2.fa  \\\n",
       "0  Densovirus_D2_non-structural_protein__1   \n",
       "1                    Densovirus_D2_ORF1__1   \n",
       "\n",
       "                          Densovirus_D3.fa  \\\n",
       "0  Densovirus_D3_non-structural_protein__1   \n",
       "1                    Densovirus_D3_ORF1__1   \n",
       "\n",
       "                          Densovirus_D4.fa  \\\n",
       "0  Densovirus_D4_non-structural_protein__1   \n",
       "1                    Densovirus_D4_ORF1__1   \n",
       "\n",
       "                          Densovirus_D5.fa  \\\n",
       "0  Densovirus_D5_non-structural_protein__1   \n",
       "1                    Densovirus_D5_ORF1__1   \n",
       "\n",
       "                          Densovirus_P1.fa  \\\n",
       "0  Densovirus_P1_non-structural_protein__1   \n",
       "1                    Densovirus_P1_ORF1__1   \n",
       "\n",
       "                          Densovirus_P2.fa  ...  \\\n",
       "0  Densovirus_P2_non-structural_protein__1  ...   \n",
       "1                    Densovirus_P2_ORF1__1  ...   \n",
       "\n",
       "                        NC_077038.1.fa  \\\n",
       "0  NC_077038.1_nonstructural_protein_2   \n",
       "1                                    *   \n",
       "\n",
       "                                 NC_077050.1.fa NC_077051.1.fa  \\\n",
       "0  NC_077050.1_putative_nonstructural_protein_1              *   \n",
       "1                                             *              *   \n",
       "\n",
       "                       NC_077065.1.fa                         NC_077069.1.fa  \\\n",
       "0  NC_077065.1_hypothetical_protein_3  NC_077069.1_nonstructural_protein_1_1   \n",
       "1  NC_077065.1_hypothetical_protein_2                                      *   \n",
       "\n",
       "                        NC_077079.1.fa                       NC_077080.1.fa  \\\n",
       "0  NC_077079.1_nonstructural_protein_1  NC_077080.1_nonstructural_protein_1   \n",
       "1                    NC_077079.1_VP1_1                    NC_077080.1_VP1_1   \n",
       "\n",
       "                        NC_077081.1.fa                       NC_077082.1.fa  \\\n",
       "0  NC_077081.1_nonstructural_protein_1  NC_077082.1_nonstructural_protein_1   \n",
       "1                                    *                    NC_077082.1_VP1_1   \n",
       "\n",
       "                        NC_077083.1.fa  \n",
       "0  NC_077083.1_nonstructural_protein_1  \n",
       "1                                    *  \n",
       "\n",
       "[2 rows x 57 columns]"
      ]
     },
     "execution_count": 73,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "df_nsp_sp"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 74,
   "metadata": {},
   "outputs": [],
   "source": [
    "df_nsp_sp_cols_to_drop = df_nsp_sp.columns[df_nsp_sp.apply(lambda col: col.astype(str).str.contains(r'\\*', regex=True).any())]"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "Examine which entries will be dropped"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 75,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "Index(['NC_001899.1.fa', 'NC_003346.1.fa', 'NC_004286.1.fa', 'NC_004287.1.fa',\n",
       "       'NC_004288.1.fa', 'NC_005041.2.fa', 'NC_005341.1.fa', 'NC_006555.1.fa',\n",
       "       'NC_011317.1.fa', 'NC_011545.2.fa', 'NC_012636.1.fa', 'NC_014357.1.fa',\n",
       "       'NC_015115.1.fa', 'NC_015718.1.fa', 'NC_018399.1.fa', 'NC_018450.1.fa',\n",
       "       'NC_019492.1.fa', 'NC_020927.1.fa', 'NC_022564.1.fa', 'NC_026943.1.fa',\n",
       "       'NC_030296.1.fa', 'NC_038376.1.fa', 'NC_040626.1.fa', 'NC_075118.1.fa',\n",
       "       'NC_077018.1.fa', 'NC_077036.1.fa', 'NC_077038.1.fa', 'NC_077050.1.fa',\n",
       "       'NC_077051.1.fa', 'NC_077069.1.fa', 'NC_077081.1.fa', 'NC_077083.1.fa'],\n",
       "      dtype='object')"
      ]
     },
     "execution_count": 75,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "df_nsp_sp_cols_to_drop"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "Save a separate `.tsv` file"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 76,
   "metadata": {},
   "outputs": [],
   "source": [
    "df_nsp_sp = df_nsp_sp.drop(columns=df_nsp_sp_cols_to_drop)\n",
    "df_nsp_sp.to_csv('phylogenomics/proteinortho_output/All.tsv', sep='\\t', index=False)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "#### **Part 6.4.4. Structural protein 2**"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 77,
   "metadata": {},
   "outputs": [],
   "source": [
    "df_interesting = df_all.drop(index=[0, 1])"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 78,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/html": [
       "<div>\n",
       "<style scoped>\n",
       "    .dataframe tbody tr th:only-of-type {\n",
       "        vertical-align: middle;\n",
       "    }\n",
       "\n",
       "    .dataframe tbody tr th {\n",
       "        vertical-align: top;\n",
       "    }\n",
       "\n",
       "    .dataframe thead th {\n",
       "        text-align: right;\n",
       "    }\n",
       "</style>\n",
       "<table border=\"1\" class=\"dataframe\">\n",
       "  <thead>\n",
       "    <tr style=\"text-align: right;\">\n",
       "      <th></th>\n",
       "      <th># Species</th>\n",
       "      <th>Genes</th>\n",
       "      <th>Alg.-Conn.</th>\n",
       "      <th>Densovirus_D1.fa</th>\n",
       "      <th>Densovirus_D2.fa</th>\n",
       "      <th>Densovirus_D3.fa</th>\n",
       "      <th>Densovirus_D4.fa</th>\n",
       "      <th>Densovirus_D5.fa</th>\n",
       "      <th>Densovirus_P1.fa</th>\n",
       "      <th>Densovirus_P2.fa</th>\n",
       "      <th>...</th>\n",
       "      <th>NC_077038.1.fa</th>\n",
       "      <th>NC_077050.1.fa</th>\n",
       "      <th>NC_077051.1.fa</th>\n",
       "      <th>NC_077065.1.fa</th>\n",
       "      <th>NC_077069.1.fa</th>\n",
       "      <th>NC_077079.1.fa</th>\n",
       "      <th>NC_077080.1.fa</th>\n",
       "      <th>NC_077081.1.fa</th>\n",
       "      <th>NC_077082.1.fa</th>\n",
       "      <th>NC_077083.1.fa</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>2</th>\n",
       "      <td>12</td>\n",
       "      <td>12</td>\n",
       "      <td>1.0</td>\n",
       "      <td>Densovirus_D1_structural_protein__1</td>\n",
       "      <td>Densovirus_D2_structural_protein__1</td>\n",
       "      <td>Densovirus_D3_structural_protein__1</td>\n",
       "      <td>Densovirus_D4_structural_protein__1</td>\n",
       "      <td>Densovirus_D5_structural_protein__1</td>\n",
       "      <td>Densovirus_P1_structural_protein__1</td>\n",
       "      <td>Densovirus_P2_structural_protein__1</td>\n",
       "      <td>...</td>\n",
       "      <td>*</td>\n",
       "      <td>*</td>\n",
       "      <td>*</td>\n",
       "      <td>*</td>\n",
       "      <td>*</td>\n",
       "      <td>*</td>\n",
       "      <td>*</td>\n",
       "      <td>*</td>\n",
       "      <td>*</td>\n",
       "      <td>*</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "<p>1 rows × 57 columns</p>\n",
       "</div>"
      ],
      "text/plain": [
       "   # Species  Genes  Alg.-Conn.                     Densovirus_D1.fa  \\\n",
       "2         12     12         1.0  Densovirus_D1_structural_protein__1   \n",
       "\n",
       "                      Densovirus_D2.fa                     Densovirus_D3.fa  \\\n",
       "2  Densovirus_D2_structural_protein__1  Densovirus_D3_structural_protein__1   \n",
       "\n",
       "                      Densovirus_D4.fa                     Densovirus_D5.fa  \\\n",
       "2  Densovirus_D4_structural_protein__1  Densovirus_D5_structural_protein__1   \n",
       "\n",
       "                      Densovirus_P1.fa                     Densovirus_P2.fa  \\\n",
       "2  Densovirus_P1_structural_protein__1  Densovirus_P2_structural_protein__1   \n",
       "\n",
       "   ... NC_077038.1.fa NC_077050.1.fa NC_077051.1.fa NC_077065.1.fa  \\\n",
       "2  ...              *              *              *              *   \n",
       "\n",
       "  NC_077069.1.fa NC_077079.1.fa NC_077080.1.fa NC_077081.1.fa NC_077082.1.fa  \\\n",
       "2              *              *              *              *              *   \n",
       "\n",
       "  NC_077083.1.fa  \n",
       "2              *  \n",
       "\n",
       "[1 rows x 57 columns]"
      ]
     },
     "execution_count": 78,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "df_interesting"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "Delete every entry with `*`"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 79,
   "metadata": {},
   "outputs": [],
   "source": [
    "df_interesting_cols_to_drop = df_interesting.columns[df_interesting.apply(lambda col: col.astype(str).str.contains(r'\\*', regex=True).any())]"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "Examine which entries will be dropped"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 80,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "Index(['NC_000936.1.fa', 'NC_001899.1.fa', 'NC_003346.1.fa', 'NC_004286.1.fa',\n",
       "       'NC_004287.1.fa', 'NC_004288.1.fa', 'NC_004289.1.fa', 'NC_004290.1.fa',\n",
       "       'NC_005040.1.fa', 'NC_005041.2.fa', 'NC_005341.1.fa', 'NC_006555.1.fa',\n",
       "       'NC_011317.1.fa', 'NC_011545.2.fa', 'NC_012636.1.fa', 'NC_012685.1.fa',\n",
       "       'NC_014357.1.fa', 'NC_015115.1.fa', 'NC_015718.1.fa', 'NC_018399.1.fa',\n",
       "       'NC_018450.1.fa', 'NC_019492.1.fa', 'NC_020927.1.fa', 'NC_022564.1.fa',\n",
       "       'NC_022748.1.fa', 'NC_026943.1.fa', 'NC_030296.1.fa', 'NC_038376.1.fa',\n",
       "       'NC_040626.1.fa', 'NC_075118.1.fa', 'NC_077018.1.fa', 'NC_077036.1.fa',\n",
       "       'NC_077038.1.fa', 'NC_077050.1.fa', 'NC_077051.1.fa', 'NC_077065.1.fa',\n",
       "       'NC_077069.1.fa', 'NC_077079.1.fa', 'NC_077080.1.fa', 'NC_077081.1.fa',\n",
       "       'NC_077082.1.fa', 'NC_077083.1.fa'],\n",
       "      dtype='object')"
      ]
     },
     "execution_count": 80,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "df_interesting_cols_to_drop"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "Save a separate `.tsv` file"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 81,
   "metadata": {},
   "outputs": [],
   "source": [
    "df_interesting_cleaned = df_interesting.drop(columns=df_interesting_cols_to_drop)\n",
    "df_interesting_cleaned.to_csv('phylogenomics/proteinortho_output/interesting_gene.tsv', sep='\\t', index=False)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "### **Part 6.5. Extract orthologs sequences**"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "For this `write_family_names.R` script will be used"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 82,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "#!/usr/bin/env Rscript\n",
      "args <- commandArgs(trailingOnly=TRUE)\n",
      "\n",
      "# Read the proteinortho table\n",
      "onetoone <- read.delim(args[1])\n",
      "\n",
      "  # Create directory to store SCOs\n",
      "  dir.create(args[2]) #e.g. data/SCOs\n",
      "\n",
      "  # Write name lists to later extract\n",
      "   for (i in 4:ncol(onetoone)) {\n",
      "   writeLines(onetoone[,i], paste0(args[2], names(onetoone[i]), \".names.txt\"))\n",
      "   }\n",
      "   dir.create(args[3])  \n",
      "   # And gene lists grouped by families\n",
      "   for (i in 1:nrow(onetoone)) {\n",
      "   oo <- as.matrix(onetoone)\n",
      "   writeLines(oo[i, 4:ncol(onetoone)], paste0(args[3], \"family\", as.character(i), \".names.txt\"))\n",
      "   }"
     ]
    }
   ],
   "source": [
    "! cat scripts/write_family_names.R"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "This script will take `.tsv` files from the previous step as the input and will write gene names to a separate files"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "#### **Part 6.5.1. Run `write_family_names.R`**"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "##### **Part 6.5.1.1. Non-structural protein**"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [],
   "source": [
    "! Rscript scripts/write_family_names.R phylogenomics/proteinortho_output/NSPs.tsv phylogenomics/NSPs/ phylogenomics/NSPs_names/"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "##### **Part 6.5.1.2. Structural protein 1**"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [],
   "source": [
    "! Rscript scripts/write_family_names.R phylogenomics/proteinortho_output/SPs.tsv phylogenomics/SPs/ phylogenomics/SPs_names/"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "##### **Part 6.5.1.3. Non-structural protein + sctructural protein 1**"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [],
   "source": [
    "! Rscript scripts/write_family_names.R phylogenomics/proteinortho_output/All.tsv phylogenomics/All/ phylogenomics/All_names/"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "##### **Part 6.5.1.4. Sctructural protein 2**"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [],
   "source": [
    "! Rscript scripts/write_family_names.R phylogenomics/proteinortho_output/interesting_gene.tsv phylogenomics/interesting_gene/ phylogenomics/interesting_gene_names/"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "#### **Part 6.5.2. Extract sequences**"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "For the 1st step we must create the `phylogenomics/data/all_pep.fa` file which will contain all the amino acid sequences from all the viruses"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 83,
   "metadata": {},
   "outputs": [],
   "source": [
    "! cat phylogenomics/data/proteomes_renamed/round_4/*.fa > phylogenomics/data/all_pep.fa"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "##### **Part 6.5.2.1. Non-structural protein**"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "First we create a directory to store the sequences"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 84,
   "metadata": {},
   "outputs": [],
   "source": [
    "! mkdir phylogenomics/data/NSPs_seqs/"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "Then we use `BBMap` to extract the sequences"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [],
   "source": [
    "%%bash\n",
    "\n",
    "for list in phylogenomics/NSPs_names/family*txt\n",
    "do bbmap/filterbyname.sh in=phylogenomics/data/all_pep.fa out=phylogenomics/data/NSPs_seqs/$(basename \"$list\" .names.txt).seq.fa include=t names=$list overwrite=true ignorejunk=true\n",
    "done"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "##### **Part 6.5.2.2. Structural protein 1**"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "First we create a directory to store the sequences"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 86,
   "metadata": {},
   "outputs": [],
   "source": [
    "! mkdir phylogenomics/data/SPs_seqs/"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "Then we use `BBMap` to extract the sequences"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [],
   "source": [
    "%%bash\n",
    "\n",
    "for list in phylogenomics/SPs_names/family*txt\n",
    "do bbmap/filterbyname.sh in=phylogenomics/data/all_pep.fa out=phylogenomics/data/SPs_seqs/$(basename \"$list\" .names.txt).seq.fa include=t names=$list overwrite=true ignorejunk=true\n",
    "done"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "##### **Part 6.5.2.3. Non-structural protein + structural protein 1**"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "First we create a directory to store the sequences"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 88,
   "metadata": {},
   "outputs": [],
   "source": [
    "! mkdir phylogenomics/data/All_seqs/"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "Then we use `BBMap` to extract the sequences"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [],
   "source": [
    "%%bash\n",
    "\n",
    "for list in phylogenomics/All_names/family*txt\n",
    "do bbmap/filterbyname.sh in=phylogenomics/data/all_pep.fa out=phylogenomics/data/All_seqs/$(basename \"$list\" .names.txt).seq.fa include=t names=$list overwrite=true ignorejunk=true\n",
    "done"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "##### **Part 6.5.2.4. Sctructural protein 2**"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "First we create a directory to store the sequences"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 90,
   "metadata": {},
   "outputs": [],
   "source": [
    "! mkdir phylogenomics/data/ig_seqs/"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "Then we use `BBMap` to extract the sequences"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [],
   "source": [
    "%%bash\n",
    "\n",
    "for list in phylogenomics/interesting_gene_names/family*txt\n",
    "do bbmap/filterbyname.sh in=phylogenomics/data/all_pep.fa out=phylogenomics/data/ig_seqs/$(basename \"$list\" .names.txt).seq.fa include=t names=$list overwrite=true ignorejunk=true\n",
    "done"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "#### **Part 6.5.3. Renaming sequences. Again.**"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "##### **Part 6.5.2.1. Non-structural protein**"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "What's wrong with seq names now? Let's take a look"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 92,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      ">Densovirus_D1_non-structural_protein__1\n"
     ]
    }
   ],
   "source": [
    "! head -1 phylogenomics/data/NSPs_seqs/family1.seq.fa"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "Well... We must leave just the viruses names"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "- Make this `>Densovirus_D1_non-structural_protein__1` look like this: `>Densovirus_D1`<br>"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "First we create a directory to store the renamed seqs"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 93,
   "metadata": {},
   "outputs": [],
   "source": [
    "! mkdir phylogenomics/data/NSPs_seqs_renamed/"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 94,
   "metadata": {},
   "outputs": [],
   "source": [
    "%%bash\n",
    "\n",
    "for file in phylogenomics/data/NSPs_seqs/*.fa\n",
    "do\n",
    "  # Process each file line by line\n",
    "  while read -r line; do\n",
    "    if [[ $line == \\>MW628494.1* ]]; then\n",
    "      # For lines starting with \">MW628494.1\", delete the first \"_\" and everything after it\n",
    "      echo \"$line\" | sed 's/_.*//' >> phylogenomics/data/NSPs_seqs_renamed/$(basename \"$file\")\n",
    "    elif [[ $line == \\>* ]]; then\n",
    "      # For all other headers, delete the second \"_\" and everything after it\n",
    "      echo \"$line\" | sed 's/\\(_[^_]*\\)\\(_.*\\)/\\1/' >> phylogenomics/data/NSPs_seqs_renamed/$(basename \"$file\")\n",
    "    else\n",
    "      # For non-header lines (sequence data), copy them as-is\n",
    "      echo \"$line\" >> phylogenomics/data/NSPs_seqs_renamed/$(basename \"$file\")\n",
    "    fi\n",
    "  done < \"$file\"\n",
    "done\n"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "Let's check if `sed` did good"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 95,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      ">Densovirus_D1\n"
     ]
    }
   ],
   "source": [
    "! head -1 phylogenomics/data/NSPs_seqs_renamed/family1.seq.fa"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "Perfect"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "##### **Part 6.5.3.2. Structural protein 1**"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "First we create a directory to store the renamed seqs"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 96,
   "metadata": {},
   "outputs": [],
   "source": [
    "! mkdir phylogenomics/data/SPs_seqs_renamed/"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 97,
   "metadata": {},
   "outputs": [],
   "source": [
    "%%bash\n",
    "\n",
    "for file in phylogenomics/data/SPs_seqs/*.fa\n",
    "do\n",
    "  # Process each file line by line\n",
    "  while read -r line; do\n",
    "    if [[ $line == \\>MW628494.1* ]]; then\n",
    "      # For lines starting with \">MW628494.1\", delete the first \"_\" and everything after it\n",
    "      echo \"$line\" | sed 's/_.*//' >> phylogenomics/data/SPs_seqs_renamed/$(basename \"$file\")\n",
    "    elif [[ $line == \\>* ]]; then\n",
    "      # For all other headers, delete the second \"_\" and everything after it\n",
    "      echo \"$line\" | sed 's/\\(_[^_]*\\)\\(_.*\\)/\\1/' >> phylogenomics/data/SPs_seqs_renamed/$(basename \"$file\")\n",
    "    else\n",
    "      # For non-header lines (sequence data), copy them as-is\n",
    "      echo \"$line\" >> phylogenomics/data/SPs_seqs_renamed/$(basename \"$file\")\n",
    "    fi\n",
    "  done < \"$file\"\n",
    "done\n"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "Let's check if `sed` did good"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 98,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      ">Densovirus_D1\n"
     ]
    }
   ],
   "source": [
    "! head -1 phylogenomics/data/SPs_seqs_renamed/family1.seq.fa"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "Perfect"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "##### **Part 6.5.3.3. Non-structural protein + structural protein 1**"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "First we create a directory to store the renamed seqs"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 99,
   "metadata": {},
   "outputs": [],
   "source": [
    "! mkdir phylogenomics/data/All_seqs_renamed/"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 100,
   "metadata": {},
   "outputs": [],
   "source": [
    "%%bash\n",
    "\n",
    "for file in phylogenomics/data/All_seqs/*.fa\n",
    "do\n",
    "  # Process each file line by line\n",
    "  while read -r line; do\n",
    "    if [[ $line == \\>MW628494.1* ]]; then\n",
    "      # For lines starting with \">MW628494.1\", delete the first \"_\" and everything after it\n",
    "      echo \"$line\" | sed 's/_.*//' >> phylogenomics/data/All_seqs_renamed/$(basename \"$file\")\n",
    "    elif [[ $line == \\>* ]]; then\n",
    "      # For all other headers, delete the second \"_\" and everything after it\n",
    "      echo \"$line\" | sed 's/\\(_[^_]*\\)\\(_.*\\)/\\1/' >> phylogenomics/data/All_seqs_renamed/$(basename \"$file\")\n",
    "    else\n",
    "      # For non-header lines (sequence data), copy them as-is\n",
    "      echo \"$line\" >> phylogenomics/data/All_seqs_renamed/$(basename \"$file\")\n",
    "    fi\n",
    "  done < \"$file\"\n",
    "done\n"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "Let's check if `sed` did good"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 101,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      ">Densovirus_D1\n"
     ]
    }
   ],
   "source": [
    "! head -1 phylogenomics/data/All_seqs_renamed/family1.seq.fa"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "Perfect"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "##### **Part 6.5.3.4. Sctructural protein 2**"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "First we create a directory to store the renamed seqs"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 102,
   "metadata": {},
   "outputs": [],
   "source": [
    "! mkdir phylogenomics/data/ig_seqs_renamed/"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 103,
   "metadata": {},
   "outputs": [],
   "source": [
    "%%bash\n",
    "\n",
    "for file in phylogenomics/data/ig_seqs/*.fa\n",
    "do\n",
    "  # Process each file line by line\n",
    "  while read -r line; do\n",
    "    if [[ $line == \\>MW628494.1* ]]; then\n",
    "      # For lines starting with \">MW628494.1\", delete the first \"_\" and everything after it\n",
    "      echo \"$line\" | sed 's/_.*//' >> phylogenomics/data/ig_seqs_renamed/$(basename \"$file\")\n",
    "    elif [[ $line == \\>* ]]; then\n",
    "      # For all other headers, delete the second \"_\" and everything after it\n",
    "      echo \"$line\" | sed 's/\\(_[^_]*\\)\\(_.*\\)/\\1/' >> phylogenomics/data/ig_seqs_renamed/$(basename \"$file\")\n",
    "    else\n",
    "      # For non-header lines (sequence data), copy them as-is\n",
    "      echo \"$line\" >> phylogenomics/data/ig_seqs_renamed/$(basename \"$file\")\n",
    "    fi\n",
    "  done < \"$file\"\n",
    "done\n"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "Let's check if `sed` did good"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 104,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      ">Densovirus_D1\n"
     ]
    }
   ],
   "source": [
    "! head -1 phylogenomics/data/ig_seqs_renamed/family1.seq.fa"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "Perfect"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "### **Part 6.6. Multiple Sequences Alignment**"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "First create directories to store the alignments"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 105,
   "metadata": {},
   "outputs": [],
   "source": [
    "! mkdir phylogenomics/MSAs/\n",
    "! mkdir phylogenomics/MSAs/NSPs/\n",
    "! mkdir phylogenomics/MSAs/SPs/\n",
    "! mkdir phylogenomics/MSAs/All/\n",
    "! mkdir phylogenomics/MSAs/IG/"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "Then launch `MAFFT` to align the sequences"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [],
   "source": [
    "! for file in phylogenomics/data/NSPs_seqs_renamed/*.fa; do mafft --auto \"$file\" > phylogenomics/MSAs/NSPs/$(basename \"$file\" .fa).aln; done"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [],
   "source": [
    "! for file in phylogenomics/data/SPs_seqs_renamed/*.fa; do mafft --auto \"$file\" > phylogenomics/MSAs/SPs/$(basename \"$file\" .fa).aln; done"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [],
   "source": [
    "! for file in phylogenomics/data/All_seqs_renamed/*.fa; do mafft --auto \"$file\" > phylogenomics/MSAs/All/$(basename \"$file\" .fa).aln; done"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [],
   "source": [
    "! for file in phylogenomics/data/ig_seqs_renamed/*.fa; do mafft --auto \"$file\" > phylogenomics/MSAs/IG/$(basename \"$file\" .fa).aln; done"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "### **Part 6.7. Trimming the alignments**"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "First create directories to store the trimmed alignments"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 110,
   "metadata": {},
   "outputs": [],
   "source": [
    "! mkdir phylogenomics/trimmed_MSAs/\n",
    "! mkdir phylogenomics/trimmed_MSAs/NSPs/\n",
    "! mkdir phylogenomics/trimmed_MSAs/SPs/\n",
    "! mkdir phylogenomics/trimmed_MSAs/All/\n",
    "! mkdir phylogenomics/trimmed_MSAs/IG/"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "Then launch `trimAl` to trim the alignments"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 111,
   "metadata": {},
   "outputs": [],
   "source": [
    "! for file in phylogenomics/MSAs/NSPs/*aln; do trimal -in $file -out phylogenomics/trimmed_MSAs/NSPs/trimmed_$(basename \"$file\") -automated1; done"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 112,
   "metadata": {},
   "outputs": [],
   "source": [
    "! for file in phylogenomics/MSAs/SPs/*aln; do trimal -in $file -out phylogenomics/trimmed_MSAs/SPs/trimmed_$(basename \"$file\") -automated1; done"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 113,
   "metadata": {},
   "outputs": [],
   "source": [
    "! for file in phylogenomics/MSAs/All/*aln; do trimal -in $file -out phylogenomics/trimmed_MSAs/All/trimmed_$(basename \"$file\") -automated1; done"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 114,
   "metadata": {},
   "outputs": [],
   "source": [
    "! for file in phylogenomics/MSAs/IG/*aln; do trimal -in $file -out phylogenomics/trimmed_MSAs/IG/trimmed_$(basename \"$file\") -automated1; done"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "### **Part 6.8. Launching `ModelFinder` to get the best substitution model**"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "First create directories to store `ModelFinder` log files"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 115,
   "metadata": {},
   "outputs": [],
   "source": [
    "! mkdir phylogenomics/model-finder/\n",
    "! mkdir phylogenomics/model-finder/NSPs/\n",
    "! mkdir phylogenomics/model-finder/SPs/\n",
    "! mkdir phylogenomics/model-finder/All/"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "Then launch `ModelFinder` to find the most appropriate substitution model"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [],
   "source": [
    "! iqtree2 -m MFP -s phylogenomics/trimmed_MSAs/NSPs/ --prefix phylogenomics/model-finder/NSPs/NSPs -T AUTO"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 117,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Best-fit model according to BIC: Q.pfam+F+I+G4\n",
      "\n",
      "List of models sorted by BIC scores: \n",
      "\n",
      "Model                  LogL         AIC      w-AIC        AICc     w-AICc         BIC      w-BIC\n",
      "Q.pfam+F+I+G4    -19698.351   39584.702 - 5.61e-05   39643.841 -  0.00241   39959.192 +    0.543\n"
     ]
    }
   ],
   "source": [
    "! head -42 phylogenomics/model-finder/NSPs/NSPs.iqtree | tail -6"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [],
   "source": [
    "! iqtree2 -m MFP -s phylogenomics/trimmed_MSAs/SPs/ --prefix phylogenomics/model-finder/SPs/SPs -T AUTO"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 119,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Best-fit model according to BIC: VT+I+G4\n",
      "\n",
      "List of models sorted by BIC scores: \n",
      "\n",
      "Model                  LogL         AIC      w-AIC        AICc     w-AICc         BIC      w-BIC\n",
      "VT+I+G4           -9482.732   19019.465 - 3.91e-13   19023.342 - 1.91e-11   19128.423 +    0.764\n"
     ]
    }
   ],
   "source": [
    "! head -42 phylogenomics/model-finder/SPs/SPs.iqtree | tail -6"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [],
   "source": [
    "! iqtree2 -m MFP -s phylogenomics/trimmed_MSAs/All/ --prefix phylogenomics/model-finder/All/All -T AUTO"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 121,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Best-fit model according to BIC: Q.pfam+I+G4\n",
      "\n",
      "List of models sorted by BIC scores: \n",
      "\n",
      "Model                  LogL         AIC      w-AIC        AICc     w-AICc         BIC      w-BIC\n",
      "Q.pfam+I+G4      -18316.088   36690.176 - 1.19e-19   36692.349 - 8.15e-19   36827.133 +    0.784\n"
     ]
    }
   ],
   "source": [
    "! head -42 phylogenomics/model-finder/All/All.iqtree | tail -6"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "### **Part 6.9. Reconstructing the phylogeny**"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "First create directories to store `IQ-TREE 2` files"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 122,
   "metadata": {},
   "outputs": [],
   "source": [
    "! mkdir phylogenomics/iqtree/\n",
    "! mkdir phylogenomics/iqtree/NSPs/\n",
    "! mkdir phylogenomics/iqtree/SPs/\n",
    "! mkdir phylogenomics/iqtree/All/"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "Then launch `IQ-TREE 2` to reconstruct the phylogeny"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [],
   "source": [
    "! iqtree2 -s phylogenomics/trimmed_MSAs/NSPs/ -m Q.pfam+F+I+G4 --prefix phylogenomics/iqtree/NSPs/NSPs -bb 1000 -T AUTO"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [],
   "source": [
    "! iqtree2 -s phylogenomics/trimmed_MSAs/SPs/ -m VT+I+G4 --prefix phylogenomics/iqtree/SPs/SPs -bb 1000 -T AUTO"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [],
   "source": [
    "! iqtree2 -s phylogenomics/trimmed_MSAs/All/ -m Q.pfam+I+G4 --prefix phylogenomics/iqtree/All/All -bb 1000 -T AUTO"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "### **Part 6.9. Fetching metadata for further visualizing in `ggtree`**"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "To make the perfect visualization of phylogenetic tree we must annotate the tree. For that we need metadata!"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "Install `phyloki`"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [],
   "source": [
    "! wget https://raw.githubusercontent.com/iliapopov17/phyloki/main/phyloki.py"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "Import `phyloki`"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 127,
   "metadata": {},
   "outputs": [],
   "source": [
    "from phyloki import *"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "Create a directory to store metadata"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 128,
   "metadata": {},
   "outputs": [],
   "source": [
    "! mkdir phylogenomics/metadata/"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "Fetch the 1st ver. of metadata"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 129,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "The request has been fulfilled.\n",
      "File saved to phylogenomics/metadata/raw_metadata.tsv\n"
     ]
    }
   ],
   "source": [
    "fetch_metadata('iljapopov17@gmail.com', 'phylogenomics/data/accession_numbers.txt', 'phylogenomics/metadata/raw_metadata.tsv')"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "Using function below we will clean `Year` column to leave only year, without the full date"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 130,
   "metadata": {},
   "outputs": [],
   "source": [
    "def clean_year(input_file, output_file):\n",
    "    \"\"\"\n",
    "    Clean the 'Year' column in a metadata .tsv file to extract only the last 4 digits.\n",
    "\n",
    "    Args:\n",
    "        input_file (str): Path to the input .tsv file.\n",
    "        output_file (str): Path to save the cleaned .tsv file.\n",
    "    \"\"\"\n",
    "    # Load the .tsv file into a DataFrame\n",
    "    df = pd.read_csv(input_file, sep=\"\\t\")\n",
    "\n",
    "    # Extract the last 4 digits of the 'Year' column\n",
    "    df['Year'] = df['Year'].apply(lambda x: str(x)[-4:] if pd.notnull(x) else 'ND')\n",
    "\n",
    "    # Save the updated DataFrame to a new .tsv file\n",
    "    df.to_csv(output_file, sep=\"\\t\", index=False)\n",
    "\n",
    "    print(f\"The 'Year' column has been cleaned.\\nFile saved to {output_file}\")"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "Run `clean_year()` function"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 131,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "The 'Year' column has been cleaned.\n",
      "File saved to phylogenomics/metadata/medium_metadata.tsv\n"
     ]
    }
   ],
   "source": [
    "clean_year('phylogenomics/metadata/raw_metadata.tsv', 'phylogenomics/metadata/medium_metadata.tsv')"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "Some of the viruses has `ND` in `Host` column"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 132,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/html": [
       "<div>\n",
       "<style scoped>\n",
       "    .dataframe tbody tr th:only-of-type {\n",
       "        vertical-align: middle;\n",
       "    }\n",
       "\n",
       "    .dataframe tbody tr th {\n",
       "        vertical-align: top;\n",
       "    }\n",
       "\n",
       "    .dataframe thead th {\n",
       "        text-align: right;\n",
       "    }\n",
       "</style>\n",
       "<table border=\"1\" class=\"dataframe\">\n",
       "  <thead>\n",
       "    <tr style=\"text-align: right;\">\n",
       "      <th></th>\n",
       "      <th>Name</th>\n",
       "      <th>Full.Name</th>\n",
       "      <th>Country</th>\n",
       "      <th>Year</th>\n",
       "      <th>Host</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>13</th>\n",
       "      <td>NC_005040.1</td>\n",
       "      <td>NC_005040.1 Myzus persicae densovirus 1</td>\n",
       "      <td>ND</td>\n",
       "      <td>ND</td>\n",
       "      <td>ND</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>15</th>\n",
       "      <td>NC_038376.1</td>\n",
       "      <td>NC_038376.1 Bombyx mori densovirus Zhenjiang</td>\n",
       "      <td>China: Zhenjiang</td>\n",
       "      <td>ND</td>\n",
       "      <td>ND</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>16</th>\n",
       "      <td>NC_003346.1</td>\n",
       "      <td>NC_003346.1 Bombyx mori densovirus 1</td>\n",
       "      <td>ND</td>\n",
       "      <td>ND</td>\n",
       "      <td>ND</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>22</th>\n",
       "      <td>NC_020927.1</td>\n",
       "      <td>NC_020927.1 Bombyx mori densovirus 3</td>\n",
       "      <td>China</td>\n",
       "      <td>ND</td>\n",
       "      <td>ND</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>24</th>\n",
       "      <td>NC_018450.1</td>\n",
       "      <td>NC_018450.1 Papilio polyxenes densovirus</td>\n",
       "      <td>Canada</td>\n",
       "      <td>2011</td>\n",
       "      <td>ND</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>25</th>\n",
       "      <td>NC_018399.1</td>\n",
       "      <td>NC_018399.1 Sibine fusca densovirus</td>\n",
       "      <td>Colombia</td>\n",
       "      <td>1977</td>\n",
       "      <td>ND</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>27</th>\n",
       "      <td>NC_005041.2</td>\n",
       "      <td>NC_005041.2 Blattella germanica densovirus 1</td>\n",
       "      <td>ND</td>\n",
       "      <td>ND</td>\n",
       "      <td>ND</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>28</th>\n",
       "      <td>NC_015115.1</td>\n",
       "      <td>NC_015115.1 Mosquito densovirus BR/07</td>\n",
       "      <td>Brazil</td>\n",
       "      <td>ND</td>\n",
       "      <td>ND</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>33</th>\n",
       "      <td>NC_011317.1</td>\n",
       "      <td>NC_011317.1 Anopheles gambiae densovirus</td>\n",
       "      <td>ND</td>\n",
       "      <td>ND</td>\n",
       "      <td>ND</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>34</th>\n",
       "      <td>NC_006555.1</td>\n",
       "      <td>NC_006555.1 Dendrolimus punctatus densovirus</td>\n",
       "      <td>China</td>\n",
       "      <td>ND</td>\n",
       "      <td>ND</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>35</th>\n",
       "      <td>NC_005341.1</td>\n",
       "      <td>NC_005341.1 Mythimna loreyi densovirus</td>\n",
       "      <td>ND</td>\n",
       "      <td>ND</td>\n",
       "      <td>ND</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>36</th>\n",
       "      <td>NC_004290.1</td>\n",
       "      <td>NC_004290.1 Acheta domestica densovirus</td>\n",
       "      <td>ND</td>\n",
       "      <td>ND</td>\n",
       "      <td>ND</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>37</th>\n",
       "      <td>NC_004288.1</td>\n",
       "      <td>NC_004288.1 Casphalia extranea densovirus</td>\n",
       "      <td>ND</td>\n",
       "      <td>ND</td>\n",
       "      <td>ND</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>39</th>\n",
       "      <td>NC_004287.1</td>\n",
       "      <td>NC_004287.1 Bombyx mori densovirus 5</td>\n",
       "      <td>ND</td>\n",
       "      <td>ND</td>\n",
       "      <td>ND</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>40</th>\n",
       "      <td>NC_004286.1</td>\n",
       "      <td>NC_004286.1 Galleria mellonella densovirus</td>\n",
       "      <td>ND</td>\n",
       "      <td>ND</td>\n",
       "      <td>ND</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>41</th>\n",
       "      <td>NC_004284.1</td>\n",
       "      <td>NC_004284.1 Junonia coenia densovirus</td>\n",
       "      <td>ND</td>\n",
       "      <td>ND</td>\n",
       "      <td>ND</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>42</th>\n",
       "      <td>NC_000936.1</td>\n",
       "      <td>NC_000936.1 Periplaneta fuliginosa densovirus</td>\n",
       "      <td>ND</td>\n",
       "      <td>ND</td>\n",
       "      <td>ND</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "           Name                                      Full.Name  \\\n",
       "13  NC_005040.1        NC_005040.1 Myzus persicae densovirus 1   \n",
       "15  NC_038376.1   NC_038376.1 Bombyx mori densovirus Zhenjiang   \n",
       "16  NC_003346.1           NC_003346.1 Bombyx mori densovirus 1   \n",
       "22  NC_020927.1           NC_020927.1 Bombyx mori densovirus 3   \n",
       "24  NC_018450.1       NC_018450.1 Papilio polyxenes densovirus   \n",
       "25  NC_018399.1            NC_018399.1 Sibine fusca densovirus   \n",
       "27  NC_005041.2   NC_005041.2 Blattella germanica densovirus 1   \n",
       "28  NC_015115.1          NC_015115.1 Mosquito densovirus BR/07   \n",
       "33  NC_011317.1       NC_011317.1 Anopheles gambiae densovirus   \n",
       "34  NC_006555.1   NC_006555.1 Dendrolimus punctatus densovirus   \n",
       "35  NC_005341.1         NC_005341.1 Mythimna loreyi densovirus   \n",
       "36  NC_004290.1        NC_004290.1 Acheta domestica densovirus   \n",
       "37  NC_004288.1      NC_004288.1 Casphalia extranea densovirus   \n",
       "39  NC_004287.1           NC_004287.1 Bombyx mori densovirus 5   \n",
       "40  NC_004286.1     NC_004286.1 Galleria mellonella densovirus   \n",
       "41  NC_004284.1          NC_004284.1 Junonia coenia densovirus   \n",
       "42  NC_000936.1  NC_000936.1 Periplaneta fuliginosa densovirus   \n",
       "\n",
       "             Country  Year Host  \n",
       "13                ND    ND   ND  \n",
       "15  China: Zhenjiang    ND   ND  \n",
       "16                ND    ND   ND  \n",
       "22             China    ND   ND  \n",
       "24            Canada  2011   ND  \n",
       "25          Colombia  1977   ND  \n",
       "27                ND    ND   ND  \n",
       "28            Brazil    ND   ND  \n",
       "33                ND    ND   ND  \n",
       "34             China    ND   ND  \n",
       "35                ND    ND   ND  \n",
       "36                ND    ND   ND  \n",
       "37                ND    ND   ND  \n",
       "39                ND    ND   ND  \n",
       "40                ND    ND   ND  \n",
       "41                ND    ND   ND  \n",
       "42                ND    ND   ND  "
      ]
     },
     "execution_count": 132,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "df = pd.read_csv('phylogenomics/metadata/medium_metadata.tsv', sep='\\t')\n",
    "nd_in_host = df[df['Host'] == 'ND']\n",
    "\n",
    "nd_in_host"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "Well.. 17 viruses from `RefSeq` has `ND` in `Host`... They are really poorly annotated. But we can predict their hosts based on their names. The host of `Periplaneta fuliginosa densovirus` must be `Periplaneta fuliginosa`. So let's do it"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "Import `re`"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 133,
   "metadata": {},
   "outputs": [],
   "source": [
    "import re"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "`extract_host()` function will assign host info based on virus name"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 134,
   "metadata": {},
   "outputs": [],
   "source": [
    "# Define a function to extract the desired host name from the Full.Name column\n",
    "def extract_host(full_name):\n",
    "    # Split the full name to remove the accession number and \"densovirus\" + everything after it\n",
    "    # This regex looks for the word 'densovirus' and removes it and any following characters.\n",
    "    return re.split(r'\\s+denso.*', full_name.split(' ', 1)[1])[0]"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "Run `extract_host()`"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 135,
   "metadata": {},
   "outputs": [],
   "source": [
    "# Update the 'Host' column where it's 'ND' with the extracted host from 'Full.Name'\n",
    "df['Host'] = df.apply(lambda row: extract_host(row['Full.Name']) if row['Host'] == 'ND' else row['Host'], axis=1)\n",
    "\n",
    "# Save the file without the index column\n",
    "df.to_csv('phylogenomics/metadata/medium_well_metadata.tsv', sep='\\t', index=False)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "Write down the information about densoviruses from this study"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 136,
   "metadata": {},
   "outputs": [],
   "source": [
    "# Step 1: Read the existing TSV file into a DataFrame\n",
    "df_existing = pd.read_csv('phylogenomics/metadata/medium_well_metadata.tsv', sep='\\t')\n",
    "\n",
    "# Step 2: Create a DataFrame for the new rows\n",
    "new_rows = {\n",
    "    'Name': ['Densovirus_D1', 'Densovirus_D2', 'Densovirus_D3', 'Densovirus_D4', 'Densovirus_D5',\n",
    "             'Densovirus_P1', 'Densovirus_P2', 'Densovirus_P3', 'Densovirus_P4', 'Densovirus_P5'],\n",
    "    'Full.Name': ['Densovirinae sp. isolate RBHC_VM_1', 'Densovirinae sp. isolate RBHC_VM_2', 'Densovirinae sp. isolate RBHC_VM_3', \n",
    "                  'Densovirinae sp. isolate RBHC_VM_4', 'Densovirinae sp. isolate RBHC_VM_5', 'Densovirinae sp. isolate RBHC_NN_1', \n",
    "                  'Densovirinae sp. isolate RBHC_NN_2', 'Densovirinae sp. isolate RBHC_NN_3', 'Densovirinae sp. isolate RBHC_NN_4', \n",
    "                  'Densovirinae sp. isolate RBHC_NN_5'],\n",
    "    'Country': ['Russia'] * 10,\n",
    "    'Year': [2024] * 10,\n",
    "    'Host': ['Bat'] * 10\n",
    "}\n",
    "\n",
    "df_new = pd.DataFrame(new_rows)\n",
    "\n",
    "# Step 3: Concatenate the new DataFrame with the existing DataFrame\n",
    "df_combined = pd.concat([df_existing, df_new], ignore_index=True)\n",
    "\n",
    "# Step 4: Save the updated DataFrame to a new TSV file (optional)\n",
    "df_combined.to_csv('phylogenomics/metadata/well_done_metadata.tsv', sep='\\t', index=False)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "### **Part 6.10. Phylogenetic trees visualization**"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "Please open `RStudio` and go through  `ggtree_journal.R` script<br>\n",
    "There are a lot of manual adjustments to the plots to make it executable"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "### **Part 6.11. Prepare SP2 gene MSA for visualization**"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 137,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Sequences renamed and saved to: phylogenomics/trimmed_MSAs/IG/trimmed_family1_renamed.seq.aln\n"
     ]
    }
   ],
   "source": [
    "rename_map = {\n",
    "    \"Densovirus_D1\": \"Densovirinae sp. isolate RBHC_VM_1\",\n",
    "    \"Densovirus_D2\": \"Densovirinae sp. isolate RBHC_VM_2\",\n",
    "    \"Densovirus_D3\": \"Densovirinae sp. isolate RBHC_VM_3\",\n",
    "    \"Densovirus_D4\": \"Densovirinae sp. isolate RBHC_VM_4\",\n",
    "    \"Densovirus_D5\": \"Densovirinae sp. isolate RBHC_VM_5\",\n",
    "    \"Densovirus_P1\": \"Densovirinae sp. isolate RBHC_NN_1\",\n",
    "    \"Densovirus_P2\": \"Densovirinae sp. isolate RBHC_NN_2\",\n",
    "    \"Densovirus_P3\": \"Densovirinae sp. isolate RBHC_NN_3\",\n",
    "    \"Densovirus_P4\": \"Densovirinae sp. isolate RBHC_NN_4\",\n",
    "    \"Densovirus_P5\": \"Densovirinae sp. isolate RBHC_NN_5\",\n",
    "    \"NC_031450.1\": \"NC_031450.1 Parus major densovirus\",\n",
    "    \"MW628494.1\": \"MW628494.1 Tenebrio molitor densovirus T2002773\"\n",
    "}\n",
    "\n",
    "input_file = \"phylogenomics/trimmed_MSAs/IG/trimmed_family1.seq.aln\"\n",
    "output_file = \"phylogenomics/trimmed_MSAs/IG/trimmed_family1_renamed.seq.aln\"\n",
    "\n",
    "with open(output_file, \"w\") as output_handle:\n",
    "    for record in SeqIO.parse(input_file, \"fasta\"):\n",
    "        if record.id in rename_map:\n",
    "            record.id = rename_map[record.id]\n",
    "            record.description = \"\"\n",
    "        SeqIO.write(record, output_handle, \"fasta\")\n",
    "\n",
    "print(\"Sequences renamed and saved to:\", output_file)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "### **Part 6.12. SP2 MSA visualization**"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "Please open `RStudio` and go through  `ggmsa_journal.R` script<br>\n",
    "There are a lot of manual adjustments to the plots to make it executable"
   ]
  }
 ],
 "metadata": {
  "kernelspec": {
   "display_name": "denso",
   "language": "python",
   "name": "python3"
  },
  "language_info": {
   "codemirror_mode": {
    "name": "ipython",
    "version": 3
   },
   "file_extension": ".py",
   "mimetype": "text/x-python",
   "name": "python",
   "nbconvert_exporter": "python",
   "pygments_lexer": "ipython3",
   "version": "3.12.3"
  }
 },
 "nbformat": 4,
 "nbformat_minor": 2
}