view imgt_loader/imgt_loader.py @ 3:beaa487ecf43 draft

Uploaded
author davidvanzessen
date Wed, 07 Dec 2016 08:36:23 -0500
parents afe85eb6572e
children
line wrap: on
line source

import pandas as pd
try:
	pd.options.mode.chained_assignment = None  # default='warn'
except:
	pass
import re
import argparse
import os

def stop_err( msg, ret=1 ):
    sys.stderr.write( msg )
    sys.exit( ret )

#docs.python.org/dev/library/argparse.html
parser = argparse.ArgumentParser()
parser.add_argument("--summ", help="The 1_Summary file from the imgt output")
parser.add_argument("--aa", help="The 5_AA-Sequence file from the imgt output")
parser.add_argument("--junction", help="The 6_Junction file from the imgt output")
parser.add_argument("--output", help="Output file")

args = parser.parse_args()

old_summary_columns = [u'Sequence ID', u'JUNCTION frame', u'V-GENE and allele', u'D-GENE and allele', u'J-GENE and allele', u'CDR1-IMGT length', u'CDR2-IMGT length', u'CDR3-IMGT length', u'Orientation']
old_sequence_columns = [u'CDR1-IMGT', u'CDR2-IMGT', u'CDR3-IMGT']
old_junction_columns = [u'JUNCTION']

added_summary_columns = [u'Functionality', u'V-REGION identity %', u'V-REGION identity nt', u'D-REGION reading frame', u'AA JUNCTION', u'Functionality comment', u'Sequence']
added_sequence_columns = [u'FR1-IMGT', u'FR2-IMGT', u'FR3-IMGT', u'CDR3-IMGT', u'JUNCTION', u'J-REGION', u'FR4-IMGT']
added_junction_columns = [u"P3'V-nt nb", u'N-REGION-nt nb', u'N1-REGION-nt nb', u"P5'D-nt nb", u"P3'D-nt nb", u'N2-REGION-nt nb', u"P5'J-nt nb", u"3'V-REGION trimmed-nt nb", 
						  u"5'D-REGION trimmed-nt nb", u"3'D-REGION trimmed-nt nb", u"5'J-REGION trimmed-nt nb", u"N-REGION", u"N1-REGION", u"N2-REGION"]

outFile = args.output

#fSummary = pd.read_csv(triplets[0][0], sep="\t", low_memory=False)
fSummary = pd.read_csv(args.summ, sep="\t", dtype=object)
#fSequence = pd.read_csv(triplets[0][1], sep="\t", low_memory=False)
fSequence = pd.read_csv(args.aa, sep="\t", dtype=object)
#fJunction = pd.read_csv(triplets[0][2], sep="\t", low_memory=False)
fJunction = pd.read_csv(args.junction, sep="\t", dtype=object)
tmp = fSummary[["Sequence ID", "JUNCTION frame", "V-GENE and allele", "D-GENE and allele", "J-GENE and allele"]]

tmp["CDR1 Seq"] = fSequence["CDR1-IMGT"]
tmp["CDR1 Length"] = fSummary["CDR1-IMGT length"]

tmp["CDR2 Seq"] = fSequence["CDR2-IMGT"]
tmp["CDR2 Length"] = fSummary["CDR2-IMGT length"]

tmp["CDR3 Seq"] = fSequence["CDR3-IMGT"]
tmp["CDR3 Length"] = fSummary["CDR3-IMGT length"]

tmp["CDR3 Seq DNA"] = fJunction["JUNCTION"]
tmp["CDR3 Length DNA"] = '1'
tmp["Strand"] = fSummary["Orientation"]
tmp["CDR3 Found How"] = 'a'

for col in added_summary_columns:
    tmp[col] = fSummary[col]

for col in added_sequence_columns:
    tmp[col] = fSequence[col]

for col in added_junction_columns:
    tmp[col] = fJunction[col]

outFrame = tmp

outFrame.columns = [u'ID', u'VDJ Frame', u'Top V Gene', u'Top D Gene', u'Top J Gene', u'CDR1 Seq', u'CDR1 Length', u'CDR2 Seq', u'CDR2 Length', u'CDR3 Seq', u'CDR3 Length', 
					u'CDR3 Seq DNA', u'CDR3 Length DNA', u'Strand', u'CDR3 Found How', u'Functionality', 'V-REGION identity %', 'V-REGION identity nt', 'D-REGION reading frame', 
					'AA JUNCTION', 'Functionality comment', 'Sequence', 'FR1-IMGT', 'FR2-IMGT', 'FR3-IMGT', 'CDR3-IMGT', 'JUNCTION', 'J-REGION', 'FR4-IMGT', 'P3V-nt nb', 
					'N-REGION-nt nb', 'N1-REGION-nt nb', 'P5D-nt nb', 'P3D-nt nb', 'N2-REGION-nt nb', 'P5J-nt nb', '3V-REGION trimmed-nt nb', '5D-REGION trimmed-nt nb', '3D-REGION trimmed-nt nb', 
					'5J-REGION trimmed-nt nb', "N-REGION", "N1-REGION", "N2-REGION"]

"""
IGHV[0-9]-[0-9ab]+-?[0-9]?D?
TRBV[0-9]{1,2}-?[0-9]?-?[123]?
IGKV[0-3]D?-[0-9]{1,2}
IGLV[0-9]-[0-9]{1,2}
TRAV[0-9]{1,2}(-[1-46])?(/DV[45678])?
TRGV[234589]
TRDV[1-3]

IGHD[0-9]-[0-9ab]+
TRBD[12]
TRDD[1-3]

IGHJ[1-6]
TRBJ[12]-[1-7]
IGKJ[1-5]
IGLJ[12367]
TRAJ[0-9]{1,2}
TRGJP?[12]
TRDJ[1-4]
"""

vPattern = [r"(IGHV[0-9]-[0-9ab]+-?[0-9]?D?)",
						r"(TRBV[0-9]{1,2}-?[0-9]?-?[123]?)",
						r"(IGKV[0-3]D?-[0-9]{1,2})",
						r"(IGLV[0-9]-[0-9]{1,2})",
						r"(TRAV[0-9]{1,2}(-[1-46])?(/DV[45678])?)",
						r"(TRGV[234589])",
						r"(TRDV[1-3])",
						r"(IGHV[0-9]S[0-9]+)"]

dPattern = [r"(IGHD[0-9]-[0-9ab]+)",
						r"(TRBD[12])",
						r"(TRDD[1-3])"]
						
jPattern = [r"(IGHJ[1-6])",
						r"(TRBJ[12]-[1-7])",
						r"(IGKJ[1-5])",
						r"(IGLJ[12367])",
						r"(TRAJ[0-9]{1,2})",
						r"(TRGJP?[12])",
						r"(TRDJ[1-4])"]

vPattern = re.compile(r"|".join(vPattern))

dPattern = re.compile(r"|".join(dPattern))

jPattern = re.compile(r"|".join(jPattern))


def filterGenes(s, pattern):
    if type(s) is not str:
        return "NA"
    res = pattern.search(s)
    if res:
        return res.group(0)
    return "NA"



outFrame["Top V Gene"] = outFrame["Top V Gene"].apply(lambda x: filterGenes(x, vPattern))
outFrame["Top D Gene"] = outFrame["Top D Gene"].apply(lambda x: filterGenes(x, dPattern))
outFrame["Top J Gene"] = outFrame["Top J Gene"].apply(lambda x: filterGenes(x, jPattern))


tmp = outFrame["VDJ Frame"]
tmp = tmp.replace("in-frame", "In-frame")
tmp = tmp.replace("null", "Out-of-frame")
tmp = tmp.replace("out-of-frame", "Out-of-frame")
outFrame["VDJ Frame"] = tmp
outFrame["CDR3 Length DNA"] = outFrame["CDR3 Seq DNA"].map(str).map(len)
safeLength = lambda x: len(x) if type(x) == str else 0
#outFrame = outFrame[(outFrame["CDR3 Seq DNA"].map(safeLength) > 0) & (outFrame["Top V Gene"] != "NA") & (outFrame["Top J Gene"] != "NA")] #filter out weird rows?
#outFrame = outFrame[(outFrame["CDR3 Seq DNA"].map(safeLength) > 0) & (outFrame["Top V Gene"] != "NA") & (outFrame["Top D Gene"] != "NA") & (outFrame["Top J Gene"] != "NA")] #filter out weird rows?
outFrame.to_csv(outFile, sep="\t", index=False, index_label="index")