metagenome

Author: gbouras13

README.md

@@ -30,7 +30,8 @@ Additionally, I would recommend reading the following guides to bacterial genome
   - [Documentation](#documentation)
   - [Method](#method)
   - [Other Features](#other-features)
-- [Quality Control](#quality-control)
+  - [Quality Control](#quality-control)
+  - [Metagenomes](#metagenomes)
   - [Installation](#installation)
     - [Conda](#conda)
     - [Pip](#pip)
@@ -152,13 +153,18 @@ Please see [here](docs/multiple_chromosomes.md) for more details and an example.
 * However, if you think you have reasonably simple metagenomic samples with long reads, in theory `plassembler` could be used.
 * We tested `plassembler` on the R10.4 sequenced ZymoBIOMICS HMW Standard sequenced in this [paper](https://www.nature.com/articles/s41592-022-01539-7), and it successfully recovered the two plasmids present in the 7 strains present in that mock community.
 
-
-# Quality Control
+## Quality Control
 
 * `plassembler` can also be used for quality control to test whether your long and short read sets come from the same isolate, even within the same species.
 
 Please see [here](docs/quality_control.md) for more details and some examples.
 
+## Metagenomes
+
+* `plassembler` is not currently recommended for metagenomic datasets, because of their high diversity, leading to difficulties in recovering chromosome-length contigs for bacteria. Additionally, Unicycler is not recommended for metagenomes. However,  `plassembler` was tested on a high depth simple mock community dataset. It worked quite nicely, but we don't anticipate it will work as well on your data!
+
+Please see [here](docs/metagenomics.md) for more details.
+
 ## Installation
 
 Plassembler has been tested on Linux and MacOS machines. 

docs/metagenomics.md

@@ -0,0 +1,33 @@
+##### Metagenomics
+
+As it currently stands, we do not recommend `plassembler` for metagenomic sequences. This is because of their high diversity, leading to difficulties in recovering chromosome-length contigs for bacteria. Additionally, Unicycler (a core dependency of Plassembler) is not recommended for metagenomes.
+
+However, we anticipate that as sequencing becomes more accurate and cheaper, it will be increasingly possible to assemble plasmids using a `plassembler` like approach from metagenomes - it's a work in progress.
+
+So as a test, we tried assembling the ZYMO HMW DNA Standard dataset from this [paper](https://www.nature.com/articles/s41592-022-01539-7), under ENA accession PRJEB48692. This mock community contains 7 bacteria and 1 fungus isolate. Notably, this dataset had extremely had deep (all bacterial chromosomes >100x coverage) and long (N50 > 20kbp) reads, so is unlikely to reflect your real-world metagenomic data as of 2023.
+
+## Get Data
+
+```
+# installation
+mamba create -n fastq-dl fastq-dl
+conda activate fastq-dl
+
+# downloads all the read sets
+fastq-dl PRJEB48692	
+
+conda deactivate
+```
+
+## Run Plassembler
+
+We decided to use `-m 10000`, because we figures that smalll plasmids would be missed by Flye anyway, and wanted complete chromosome assemblies, and a `-c 500000`. We used 32 threads on 16 cores and allocated 80 GB of RAM.
+
+```
+plassembler run -d Plassembler_DB -l ERR7287988.fastq.gz -1 ERR7255689_1.fastq.gz -2 ERR7255689_2.fastq.gz \
+-f -t 32 -q 10 -o zymo_R10.4_flye -m 10000 -c 500000
+```
+
+`plassembler` took around 8 hours (wall clock) to finish and excitingly we assembled all 7 bacterial chromosomes using Flye (unsurprising!) along with the 5 plasmids indicated in the ground truth (1 _E. coli_ 100kbp, 1 _S. enterica_ 49kbp and 3 small _S. aureus_ plasmids (6, 2 and 2 kbp)) with genome fraction 100% from QUAST. 
+
+So in theory `plassembler`  might work on metagenomes, but I would caution against using it, for now.

mkdocs.yml

@@ -27,4 +27,5 @@ nav:
     - OTHER FEATURES:
       - Quality Control: quality_control.md
       - Multiple Chromosomes: multiple_chromosome.md
+      - Metagenomes: metagenomics.md