Ecosyste.ms: Awesome

An open API service indexing awesome lists of open source software.

Awesome Lists | Featured Topics | Projects

https://github.com/ssi-dk/ecoli_pathotypes


https://github.com/ssi-dk/ecoli_pathotypes

Last synced: 3 days ago
JSON representation

Awesome Lists containing this project

README 

        
# A bunch of stuff for downloading and checking stuff on a bunch of ecoli genomes from enterobase

The scripts work together based on the assumption that stuff is being done in a particular project directory. The R scripts take this directory as a command line argument (this must end with slash, e.g. mydir/). It's also assumed at the [slurm](https://slurm.schedmd.com) scheduler is used for sending jobs on the cluster.

# Dependencies

We used [quast](http://bioinf.spbau.ru/quast) for checking basic statistics on the assemblies, [kma](https://bitbucket.org/genomicepidemiology/kma/src/master/) for checking matches to specific genes.

Python dependencies: pandas  

R dependencies: tidyverse, Biostrings, jsonlite, cowplot  

Additionally, these scripts assume we're using the slurm job manager.  

# Downloads

First, metadata containing assemblystats, cgmlst, wgmlst phylotypes and achtman mlst was downloaded off enterobase manually, in our case such that the the release date was at most 22-09-27, and named as following:  

'meta_with_assemblystats.tsv'  

'meta_with_cgmlst.tsv'  

meta_with_wgmlst.tsv  

meta_with_phylotypes.tsv  

wrangle_ecoli_clean.R was used for combining these metadata files and filtering them to get a list of desired samples (e.g. must have collection year, source).  

download_links.py was used for downloading assembly links.  

generate_download.R is a script that calls download_fastas.py on a subset of assembly links and does this using slurm.  

# QC and finding genes

basic_fasta_stats.py uses quast to perform basic QC of the assemblies and this is then collected using collect_quast_tsvs.py and appended to the metadata using get_final_meta.R  

The provided genes used for classifying pathotypes were indexed with kma and then send_kma_jobs.py is used which calls run_kma_with_genes_as_db.py and uses kma to search for the provided virulence genes in the assemblies:  

python send_kma_jobs.py -i fasta_files -k /path/to/kma_binary -db /path/to/indexed_genes -o concatenated_genes_kma_results -chunk 10000  

The results are concatenated with collect_kma_results.py

Then finally generate_gene_matrix.R is used to both generate a binary gene matrix for the samples but also call pathotypes depending on the presence of certain genes in the assemblies, and this uses the concatenated kma results tsv file.

In order for a gene to be called present it must have at least 90% identity and at least 90% coverage.  

For further analyses, genomes with a size lower than 4MB and number of contigs >= 800 were excluded.