GitLocker: The Coding Marketplace

Description:

pycov3 2.1.0

pycov3

A package for generating cov3 files which are generated from sam files giving coverage information and a fasta file giving binned contigs. Cov3 files are used as input for the DEMIC R package which calculates PTR, an estimate for bacterial growth rates.
Installation
PyPi
pip install pycov3
pycov3 -h

Bioconda
conda create -n pycov3 -c conda-forge -c bioconda pycov3
conda activate pycov3
pycov3 -h

DockerHub
docker pull ctbushman/pycov3:latest
docker run --rm --name pycov3 pycov3 pycov3 -h

GitHub
git clone https://github.com/Ulthran/pycov3.git
cd pycov3/
pip install .
pycov3 -h

Usage
Use -h to see options for running the CLI.
$ pycov3 -h

The FASTAs should all be in one directory with names of the format {sample}.{bin_name}.fasta/.fa/.fna and the SAMs should also all be in one directory with names of the format {sample}_{bin_name}.sam. The output COV3 files will be written to a directory with names of the format {sample}.{bin_name}.cov3.
You can also use the library in your own code. Create a SAM directory and FASTA directory, set any non-default window or coverage parameters, then create a COV3 directory and use it to generate a COV3 file for each contig set in the FASTA directory.
from pycov3.Directory import Cov3Dir, FastaDir, SamDir

sam_d = SamDir(Path("/path/to/sams/"), False)

window_params = {
"window_size": None,
"window_step": None,
"edge_length": sam_d.calculate_edge_length(),
}
coverage_params = {
"mapq_cutoff": None,
"mapl_cutoff": None,
"max_mismatch_ratio": None,
}
window_params = {k: v for k, v in window_params.items() if v is not None}
coverage_params = {k: v for k, v in coverage_params.items() if v is not None}

fasta_d = FastaDir(Path("/path/to/fastas/"), False)

cov3_d = Cov3Dir(
Path(args.out_dir),
False,
fasta_d.get_filenames(),
window_params,
coverage_params,
)

cov3_d.generate(sam_d, fasta_d)

Alternatively, to use the bare application logic and do all the file handling yourself, you can use the Cov3Generator class which takes a list of generators as SAM inputs and a generator as a FASTA input.
from pycov3.Cov3Generator import Cov3Generator
from pycov3.File import Cov3File

cov3_generator = Cov3Generator(
sam_generators,
fasta_generator,
sample,
bin_name,
window_params,
**coverage_params,
)

cov3_dict = cov3_generator.generate_cov3()

# Write output
cov3_file = Cov3File(Path(/path/to/output/), "001")
cov3_file.write_generator(cov3_generator.generate_cov3())

Resource Requirements
Threads: pycov3 uses multiprocessing to parallelize processing of input fastas. Increasing --thread_num up to the number of input fastas should improve runtime, with no benefits beyond that number.
Memory: pycov3 uses generators as much as possible. The main memory users are the Contig objects, which each hold a contig's sequence and information for each Window over its length. There is also a coverages dictionary that could potentially grow to the size of the largest contig (but that is very unlikely). At a minimum, twice the size of the largest contig should be given per thread.
Algorithmic Complexity: Assuming enough threads are provided to have each fasta file processed separately, the time complexity is roughly O(cwsr).
c: Number of contigs in fasta
s: Number of sam files
w: Max number of windows per contig
r: Max number of records per sam file
Help
Please use the Issues on this repo for any problems, questions, or suggestions.