hamburger

HMmer Based UndeRstandinG of gene clustERs

A tool to extract and analyse contiguous sets of genes in bacterial genomes, given a concatenated set of protein hidden markov models (HMMs). Written in python 3 and R (for T6SS filtering/grouping)

Hamburger uses sets of concatenated HMMs to search for sets of genes co-localised in a genome, using gff3 files as input. Input gff3 files need both the annotation, and fasta sequence (preceded by ##FASTA in the file, as produced by prokka)

Hamburger can be used as follows - where < > denotes user specific addresses/input:

Firstly cloning the git repository using git clone (into desired local directory)

git clone https://www.github.com/djw533/hamburger

Then running hamburger to search using a set of HMMs concatenated into a single file

python <hamburger_directory>/scripts/hamburger.py -g *.gff -i <concatenated set of hmms.hmm> -o <new output directory to write in> -m <minimum number of genes needed to report gene cluster> -l <max permitted gap of non-model genes between genes of interest>

An automated search and subtyping of T6SSs can be used as follows:

python <hamburger_directory>/scripts/hamburger.py -g *.gff -t  -o <new output directory to write in>

Followed by a filtering step

python  <hamburger_directory>/scripts/post_hamburger_t6SS_search.py -i <output directory used for hamburger search>

Output for loading into itol can be requested with the -q flag

The following dependencies are required:

System:

HMMER
Python 3

(and for T6SS subtyping):

Muscle
Fasttree
R

Python libraries:

Biopython
tqdm
pandas

R libraries (for T6SS subtyping) - can be installed using the install_R_packages.R script:

ggplot2
dplyr
gggenes
RColorBrewer
castor
ape
ggtree
glue

usage: hamburger.py [-h] [-i MANDATORY] [-a ACCESSORY] [-g GFF [GFF ...]] [-f FASTA [FASTA ...]]
                    [-m MIN_GENES] [-l GENES_GAP] [-u UPSTREAM] [-d DOWNSTREAM] [-c CUTOFF] [-t]
                    [-n NUM_THREADS] [-k] [-o OUTPUT] [-q] [-w] [-s] [-v]

Extract and plot gene_clusters based on hmm profiles
--------------------HaMBURGER--------------------

-------HMmer Based UndeRstandinG of gene clustERs------

              _....----'''----...._
           .-'  o    o    o    o   '-.
          /  o    o     o    o   o    \  	
       __/__o___o_ _ o___ _o _ o_ _ _o_\__
      /                                   \ 	
      \___________________________________/
        \~`-`.__.`-~`._.~`-`~.-~.__.~`-`/
         \                             /
          `-._______________________.-'


optional arguments:
  -h, --help            show this help message and exit
  -i MANDATORY, --mandatory MANDATORY
                        Mandatory hmm profile input <required if not using -t flag>
  -a ACCESSORY, --accessory ACCESSORY
                        Accessory hmm profile input
  -g GFF [GFF ...], --gff GFF [GFF ...]
                        Gff file(s) to search. Can be used in combination with --fasta or standalone (if fasta is appended to the end of the gff inbetween a line with ##FASTA)
  -f FASTA [FASTA ...], --fasta FASTA [FASTA ...]
                        Fasta file(s) to search. Can be used in combination with --gff or standalone, in which case prodigal will predict CDSs
  -m MIN_GENES, --min_genes MIN_GENES
                        Minimum number of genes in gene cluster, default = 4
  -l GENES_GAP, --genes_gap GENES_GAP
                        Maximum number of genes gap between hits, default = 10
  -u UPSTREAM, --upstream UPSTREAM
                        Number of nucleotides to include upstream/"right" of gene cluster, default = 0
  -d DOWNSTREAM, --downstream DOWNSTREAM
                        Number of nucleotides to include downstream/"left" of gene cluster, default = 0
  -c CUTOFF, --cutoff CUTOFF
                        Cutoff HMMER score for each hit, default = 20
  -t, --t6ss            Automatic searching for T6SSs, uses min_genes = 8, genes_gap = 12, mandatory hmm profile of all 13 tss genes
  -n NUM_THREADS, --num_threads NUM_THREADS
                        Number of threads to use, default = 1
  -k, --keep_files      Keep all intermediate files produced, default = False
  -o OUTPUT, --output OUTPUT
                        Output directory, default =Hamburger output. Will not write over a previously existing output folder!
  -q, --itol            Create itol output for number of T6SSs and subtypes per strain
  -w, --overwrite       Overwrite existing blast database directory
  -s, --save_gffs       Save output gff files
  -v, --version         Print version and exit