avirfrog/fasta-parser

A lightweight Python module to read and write FASTA sequence records

fasta-parser

A lightweight Python module to read and write sequence records in FASTA format.

The design was inspired by the utility of BioPython’s SeqIO, which supports many sequence formats. This repo focuses only on FASTA records. It is faster than BioPython and can handle compressed FASTA files (gz, bz2, zip, zstandard).

Requirements

Python >= 3.8

Optional dependencies

Two dependencies can be installed later if needed:

1. lz4 >= 4.0.2

   Install with: pip install lz4

2. zstandard >= 0.18.0

   Install with pip install zstandard.

Quick Start

Typical usage is to read a FASTA file and loop over the record(s).

import fasta

for record in fasta.parse('test/test.fasta'):
    print(record.id, len(record), record.seq[:10], record.desc)

Output:

NP_002433.1  362   METDAPQPGL   RNA-binding protein Musashi homolog 1 [Homo sapiens]
ENO94161.1    79   MKLLISGLGP   RRM domain-containing RNA-binding protein
sequence     292   MKLSKIALMM

Usage

This module contains the Record class representing a FASTA sequence record and the parse() function to read FASTA records from a file.

Record object

Record is an object that contains information on a FASTA sequence record, including id, description, and the sequence itself.

import fasta

record = fasta.Record(
    id='NP_950171.2', 
    seq='MEEEAETEEQQRFSYQQRLKAAVHYTVGCLCEEVALDKEMQFSKQTIAAISELTFRQCENFAKDLEMFASICRKRQE',
    desc='APITD1-CORT protein isoform 2 [Homo sapiens]'
)

print(record.id)            # NP_950171.2
print(record.desc)          # APITD1-CORT protein isoform 2 [Homo sapiens]
print(record.seq)           # MEEEAE..
print(record.description)   # >NP_950171.2 G APITD1-CORT protein isoform 2 [Homo sapiens]
print(len(record))          # 77
print('EEEA' in record)     # True

By default, the sequence line in the FASTA format is wrapped to 70 characters. You can provide the line length. Use zero (or None) for no wrapping.

print(record)
# >NP_950171.2 APITD1-CORT protein isoform 2 [Homo sapiens]
# MEEEAETEEQQRFSYQQRLKAAVHYTVGCLCEEVALDKEMQFSKQTIAAISELTFRQCENFAKDLEMFAS
# ICRKRQE

print(record.format(wrap=30))
# >NP_001382951.1 G protein subunit gamma 5 [Homo sapiens]
# MEEEAETEEQQRFSYQQRLKAAVHYTVGCL
# CEEVALDKEMQFSKQTIAAISELTFRQCEN
# FAKDLEMFASICRKRQE

print(record.format(wrap=None))
# >NP_950171.2 APITD1-CORT protein isoform 2 [Homo sapiens]
# MEEEAETEEQQRFSYQQRLKAAVHYTVGCLCEEVALDKEMQFSKQTIAAISELTFRQCENFAKDLEMFASICRKRQE

parse

The parse() function is a generator to read FASTA records as Record objects one by one from a file (plain FASTA or compressed using gzip or bzip2). Because only one record is created at a time, very little memory is required.

import fasta

for record in fasta.parse('test/test.fasta.gz'):
    print(record.id)

For some tasks you may need to have a reusable access to the records. For this purpose, you can use the built-in Python list() function to turn the iterator into a list:

import fasta

records = list(fasta.parse('test/test.fasta.gz'))
print(records[0].id)   # First record
print(records[-1].id)  # Last record

Another common task is to index your records by sequence identifier. Use to_dict() to turn a Record iterator (or list) into a dictionary.

import fasta

records = fasta.to_dict(fasta.parse('test/test.fasta.gz'))
print(records['NP_002433.1'])   # Use any record id

Test

You can run tests to ensure that the module works as expected.

./test.py

License

GNU General Public License, version 3