vcfparser

Python (version >=3.6) package for parsing the genomics and transcriptomics VCF data.

• Free software: MIT license
• Documentation: https://vcfparser.readthedocs.io.

Features

• No external dependency except python (version >=3.6).
• Minimalistic in nature.
• Provides a lot of features to API users.
• Cython compiling is provided to optimize performance.

Installation

Method A:

VCFsimplify uses vcfparser API, so the package is readily available if VCFsimplify is already installed.

This is only preferred while developing/optimizing VcfSimplify along with vcfparser.

Navigate to the VCFsimplify directory -> activate python -> call the ‘vcfparser’ package.

$ C:\Users\>cd VCF-Simplify
$ C:\Users\>cd VCF-Simplify>dir
  Volume in drive C is StorageDrive
  Volume Serial Number is .........

  Directory of C:\Users\VCF-Simplify

  07/12/2020  10:14 AM    <DIR>          .
  07/12/2020  10:14 AM    <DIR>          ..
  07/12/2020  08:55 AM    <DIR>          .github
  ............................
  ............................
  07/12/2020  10:42 AM    <DIR>          vcfparser
  07/12/2020  08:55 AM             1,494 VcfSimplify.py
          11 File(s)     20,873,992 bytes
          13 Dir(s)  241,211,793,408 bytes free

$ C:\Users\VCF-Simplify>python
Python 3.8.1 (tags/v3.8.1:1b293b6, Dec 18 2019, 22:39:24) [MSC v.1916 (Intel)] on win32
Type "help", "copyright", "credits" or "license" for more information.
>>> from vcfparser import VcfParser
>>>

Method B (preferred method):

Pip is the preferred method of installing and using vcfparser API if custom python scripts/app are being developed.

$ pip install vcfparser

Method C:

For offline install, or in order to build from the source code, follow advance install.

Cythonize (optional but helpful)

The installed “vcfparser” package can be cythonized to optimize performance. Cythonizing the package can increase the speed of the parser by about x.x - y.y (?) times.

TODO: Bhuwan - add required cython method in here

Usage

>>> from vcfparser import VcfParser
>>> vcf_obj = VcfParser('input_test.vcf')

Get metadata information from the vcf file

>>> metainfo = vcf_obj.parse_metadata()
>>> metainfo.fileformat
'VCFv4.2'
>>> metainfo.filters_
[{'ID': 'LowQual', 'Description': 'Low quality'}, {'ID': 'my_indel_filter', 'Description': 'QD < 2.0 || FS > 200.0 || ReadPosRankSum < -20.0'}, {'ID': 'my_snp_filter', 'Description': 'QD < 2.0 || FS > 60.0 || MQ < 40.0 || MQRankSum < -12.5 || ReadPosRankSum < -8.0'}]
>>> metainfo.alt_
[{'ID': 'NON_REF', 'Description': 'Represents any possible alternative allele at this location'}]
>>> metainfo.sample_names
['ms01e', 'ms02g', 'ms03g', 'ms04h', 'MA611', 'MA605', 'MA622']
>>> metainfo.record_keys
['CHROM', 'POS', 'ID', 'REF', 'ALT', 'QUAL', 'FILTER', 'INFO', 'FORMAT', 'ms01e', 'ms02g', 'ms03g', 'ms04h', 'MA611', 'MA605', 'MA622']

Get Records from the vcf file

>>> records = vcf_obj.parse_records()
    # Note: Records are returned as generator.
>>> first_record = next(records)
>>> first_record.CHROM
'2'
>>> first_record.POS
'15881018'
>>> first_record.REF
'G'
>>> first_record.ALT
'A,C'
>>> first_record.QUAL
'5082.45'
>>> first_record.FILTER
['PASS']
>>> first_record.get_mapped_samples()
{'ms01e': {'GT': './.', 'PI': '.', 'GQ': '.', 'PG': './.', 'PM': '.', 'PW': './.', 'AD': '0,0', 'PL': '0,0,0,.,.,.', 'DP': '0', 'PB': '.', 'PC': '.'}, 'ms02g': {'GT': './.', 'PI': '.', 'GQ': '.', 'PG': './.', 'PM': '.', 'PW': './.', 'AD': '0,0', 'PL': '0,0,0,.,.,.', 'DP': '0', 'PB': '.', 'PC': '.'}, 'ms03g': {'GT': './.', 'PI': '.', 'GQ': '.', 'PG': './.', 'PM': '.', 'PW': './.', 'AD': '0,0', 'PL': '0,0,0,.,.,.', 'DP': '0', 'PB': '.', 'PC': '.'}, 'ms04h': {'GT': '1/1', 'PI': '.', 'GQ': '6', 'PG': '1/1', 'PM': '.', 'PW': '1/1', 'AD': '0,2', 'PL': '49,6,0,.,.,.', 'DP': '2', 'PB': '.', 'PC': '.'}, 'MA611': {'GT': '0/0', 'PI': '.', 'GQ': '78', 'PG': '0/0', 'PM': '.', 'PW': '0/0', 'AD': '29,0,0', 'PL': '0,78,1170,78,1170,1170', 'DP': '29', 'PB': '.', 'PC': '.'}, 'MA605': {'GT': '0/0', 'PI': '.', 'GQ': '9', 'PG': '0/0', 'PM': '.', 'PW': '0/0', 'AD': '3,0,0', 'PL': '0,9,112,9,112,112', 'DP': '3', 'PB': '.', 'PC': '.'}, 'MA622': {'GT': '0/0', 'PI': '.', 'GQ': '99', 'PG': '0/0', 'PM': '.', 'PW': '0/0', 'AD': '40,0,0', 'PL': '0,105,1575,105,1575,1575', 'DP': '40', 'PB': '.', 'PC': '.\n'}}

TODO: Bhuwan (priority - high) The very last example “first_record.get_mapped_samples()” is returning the value of the last sample/key with “n”. i.e: ‘PC’: ‘.n’ Please fix that issue - strip(‘n’) in the line before parsing.

Alternately, we can loop over each record by using a for-loop:

for record in records:
    chrom = record.CHROM
    pos = record.POS
    id = record.ID
    ref = record.REF
    alt = record.ALT
    qual = record.QUAL
    filter = record.FILTER
    format_ = record.format_
    infos = record.get_info_dict()
    mapped_sample = record.get_mapped_samples()

For more specific use cases please check the examples in the following section:

For tutorials in metadata, please follow Metadata Tutorial.

For tutorials in record parser, please follow Record Parser Tutorial.