# Extraction of the pbp TPD domains from Strep pneumo gffs #

## Installation ##

This requires conda, please install conda first [here](https://docs.conda.io/projects/conda/en/latest/user-guide/install/)
Once installed clone the repo:

`git clone https://github.com/jdaeth274/pbp_tpd_extraction`

Then use the environment.yml file to install the dependencies with conda

`conda env create --file=environment.yml`

Activate this environment using 

`conda activate pbp_tpd_env`

## Usage ##

### Penicillin resistance ###

For pbp extraction, once in the pbp_tpd_env created from above use the following script command
to run and extract a csv of the isolate name and the pbp resistance category. 

`bash ./bash/pbp_gene_extraction.sh gff_list.txt fasta_list.txt out_csv_name.csv`

Here the gff_list is a txt file with the location of a gff file on each line.
The fasta list is complementary to this, with the same fasta file on each line, 
as corresponds to the gff list.

### Co-trimoxazole resistance ###

This method relies on HMMs to find the _folP_ and _dhfR_ genes within sequences, 
looking for these genes to find the resistance mutations. Each gene needs to be run
separately. For _folP_ run:   
`python python/pen_checker_cdc.py --gff gff_list.txt --pbp folP --gene folP --tlength 2159 --output output.csv --tolerance 100 --data_dir ./data`   
   
Then for _dhfR_ run:   
`python python/pen_checker_cdc.py --gff gff_list.txt --pbp dhfR --gene dhfR --tlength 2159 --output output.csv --tolerance 100 --data_dir ./data`