def name2geneMap(organism1, taxID1, organism2, taxID2, gene_info):
"""
This function maps gene ids between two species by full name.
* organism1: string. the 1st organism. eg: 'cho'
* taxID1: taxonomy ID. eg: 10029
* organism2: string. the 2nd organism. eg: 'human'
* taxID2: taxonomy ID. eg: 9606
return: [cho2human.txt, cho_gene_info.txt]. For 'cho2human.txt' file, ID mapping based on gene names.
1st column is unique geIDs, 2nd column is all of the genes mapping to
"""
org1_lib = {}
org2_lib = {}
gene1 = []
gene2 = []
org1_file = extract_from_gene2ref(gene_info,
taxID1,
organism1,
columnNum=[2, 9])
org2_file = extract_from_gene2ref(gene_info,
taxID2,
organism2,
columnNum=[2, 9])
outputFile = org1_file[:-3] + organism1 + '2' + organism2 + '.txt'
# build libray for organism1
result = open(org1_file, 'r')
for line in result:
item = line[:-1].split('\t')
# if gene name has many ids
if item[1] in org1_lib:
org1_lib[item[1]].append(item[0])
else:
org1_lib[item[1]] = [item[0]]
gene1.append(item[1])
# build library for organism2
result = open(org2_file, 'r')
for line in result:
item = line[:-1].split('\t')
if item[1] in org2_lib:
org2_lib[item[1]].append(item[0])
else:
org2_lib[item[1]] = [item[0]]
gene2.append(item[1])
# find intersect gene name
intersectName = set(gene1).intersection(gene2)
# from name to gene id, and output
output = open(outputFile, 'w')
for name in intersectName:
geneID1 = org1_lib[name]
geneID2 = org2_lib[name]
# geneID1 and geneID2 are lists and can have many items in side it.
for fir in geneID1:
output.write(fir + '\t' + ';'.join(geneID2) + '\n')
output.close()
return [outputFile, org1_file]
def name2geneMap(organism1,taxID1,organism2,taxID2,gene_info):
"""
This function maps gene ids between two species by full name.
* organism1: string. the 1st organism. eg: 'cho'
* taxID1: taxonomy ID. eg: 10029
* organism2: string. the 2nd organism. eg: 'human'
* taxID2: taxonomy ID. eg: 9606
return: [cho2human.txt, cho_gene_info.txt]. For 'cho2human.txt' file, ID mapping based on gene names.
1st column is unique geIDs, 2nd column is all of the genes mapping to
"""
org1_lib = {}; org2_lib = {}
gene1 = []; gene2=[]
org1_file = extract_from_gene2ref(gene_info,taxID1,organism1,columnNum=[2,9])
org2_file = extract_from_gene2ref(gene_info,taxID2,organism2,columnNum=[2,9])
outputFile = org1_file[:-3] + organism1 + '2' + organism2 + '.txt'
# build libray for organism1
result = open(org1_file,'r')
for line in result:
item = line[:-1].split('\t')
# if gene name has many ids
if item[1] in org1_lib:
org1_lib[item[1]].append(item[0])
else:
org1_lib[item[1]] = [item[0]]
gene1.append(item[1])
# build library for organism2
result = open(org2_file,'r')
for line in result:
item = line[:-1].split('\t')
if item[1] in org2_lib:
org2_lib[item[1]].append(item[0])
else:
org2_lib[item[1]] = [item[0]]
gene2.append(item[1])
# find intersect gene name
intersectName = set(gene1).intersection(gene2)
# from name to gene id, and output
output = open(outputFile,'w')
for name in intersectName:
geneID1 = org1_lib[name]
geneID2 = org2_lib[name]
# geneID1 and geneID2 are lists and can have many items in side it.
for fir in geneID1:
output.write(fir + '\t' + ';'.join(geneID2) + '\n')
output.close()
return [outputFile,org1_file]
def blastp2geneMap(blastFiles,organism1,taxID1,organism2,taxID2,gene2refseq,topNum,IDtype='protein'):
"""
This function transfers blastp results into geneID mapping results
* blastFiles: a list of 2 way blastp tabular result files. eg: [blast1.txt,blast2.txt]
* organism1: string. the 1st organism. eg: 'cho'
* organism2: string. the 2nd organism. eg: 'human'
* gene2refseq: filename. eg: 'gene2refseq'
returns a list of four files. For first 2: each with two columns of gene ID mapping.
For thre rest 2: gene2refseq files for each organism
"""
# extract gene, accession, protein mapping
if IDtype == 'protein':
columnNum = [2,6,7,16]
else:
columnNum = [2,4,5,16]
org1ref = extract_from_gene2ref(gene2refseq,taxID1,organism1,columnNum)
org2ref = extract_from_gene2ref(gene2refseq,taxID2,organism2,columnNum)
result = []
switches = ['False','True']
for blast,switch in zip(blastFiles,switches):
# extract protein id map
pr_id_map = extract_blast_ID_map(blast,topNum,switch) # pr_id_map: cho2human.top1.txt
# protein id mapping to gene id mapping
gene_id_map = mRNA_prIDMap2geneIDMap(pr_id_map,org1ref,org2ref,IDtype) # cho2human.top1.gene.txt
# get unique line of mapping
uniqFile = gene_id_map[:-3] + 'uniqline.txt' # cho2human.top1.gene.uniqline.txt
interFile = gene_id_map[:-3] + 'inter.txt'
# # # get unique id mappings (each line is unique, but genes are not unique)
if switch == 'True':
cmd1 = ('awk -F $\'\\t\' \'BEGIN {OFS} {printrow}\' {input} > {output}').format(
OFS='{FS=\"\\t\"; OFS=FS}',printrow='{print $1,$2}',input=gene_id_map,output=interFile)
cmd2 = ('sort -k2,2 -k1,1 {input} | uniq > {output}').format(input=interFile,output=uniqFile)
else:
cmd1 = ('awk -F $\'\\t\' \'BEGIN {OFS} {printrow}\' {input} > {output}').format(
OFS='{FS=\"\\t\"; OFS=FS}',printrow='{print $1,$2}',input=gene_id_map,output=interFile)
cmd2 = ('sort -k1,1 -k2,2 {input} | uniq > {output}').format(input=interFile,output=uniqFile)
subprocess.call(cmd1,shell=True)
subprocess.call(cmd2,shell=True)
subprocess.call(('rm {inter}').format(inter=interFile),shell=True)
# unique gene ID in 1st column
uniq = uniq1stGene(uniqFile) # cho2human.top1.gene.uniqline.uniq1stgene.txt
result.append(uniq)
result.extend([org1ref,org2ref])
return result
def blastp2geneMap(blastFiles,
organism1,
taxID1,
organism2,
taxID2,
gene2refseq,
topNum,
IDtype='protein'):
"""
This function transfers blastp results into geneID mapping results
* blastFiles: a list of 2 way blastp tabular result files. eg: [blast1.txt,blast2.txt]
* organism1: string. the 1st organism. eg: 'cho'
* organism2: string. the 2nd organism. eg: 'human'
* gene2refseq: filename. eg: 'gene2refseq'
returns a list of four files. For first 2: each with two columns of gene ID mapping.
For thre rest 2: gene2refseq files for each organism
"""
# extract gene, accession, protein mapping
if IDtype == 'protein':
columnNum = [2, 6, 7, 16]
else:
columnNum = [2, 4, 5, 16]
org1ref = extract_from_gene2ref(gene2refseq, taxID1, organism1, columnNum)
org2ref = extract_from_gene2ref(gene2refseq, taxID2, organism2, columnNum)
result = []
switches = ['False', 'True']
for blast, switch in zip(blastFiles, switches):
# extract protein id map
pr_id_map = extract_blast_ID_map(
blast, topNum, switch) # pr_id_map: cho2human.top1.txt
# protein id mapping to gene id mapping
gene_id_map = mRNA_prIDMap2geneIDMap(pr_id_map, org1ref, org2ref,
IDtype) # cho2human.top1.gene.txt
# get unique line of mapping
uniqFile = gene_id_map[:-3] + 'uniqline.txt' # cho2human.top1.gene.uniqline.txt
interFile = gene_id_map[:-3] + 'inter.txt'
# # # get unique id mappings (each line is unique, but genes are not unique)
if switch == 'True':
cmd1 = (
'awk -F $\'\\t\' \'BEGIN {OFS} {printrow}\' {input} > {output}'
).format(OFS='{FS=\"\\t\"; OFS=FS}',
printrow='{print $1,$2}',
input=gene_id_map,
output=interFile)
cmd2 = ('sort -k2,2 -k1,1 {input} | uniq > {output}').format(
input=interFile, output=uniqFile)
else:
cmd1 = (
'awk -F $\'\\t\' \'BEGIN {OFS} {printrow}\' {input} > {output}'
).format(OFS='{FS=\"\\t\"; OFS=FS}',
printrow='{print $1,$2}',
input=gene_id_map,
output=interFile)
cmd2 = ('sort -k1,1 -k2,2 {input} | uniq > {output}').format(
input=interFile, output=uniqFile)
subprocess.call(cmd1, shell=True)
subprocess.call(cmd2, shell=True)
subprocess.call(('rm {inter}').format(inter=interFile), shell=True)
# unique gene ID in 1st column
uniq = uniq1stGene(
uniqFile) # cho2human.top1.gene.uniqline.uniq1stgene.txt
result.append(uniq)
result.extend([org1ref, org2ref])
return result
def rosetta_stone(blastFile,organism1,taxID1,organism2,taxID2,gene2refseq):
"""
This functions detects rosetta stone pairs in the blast result.
* blastFile: fileanme of tabular blast output.
"""
# # (1) change first 2 columns into ID, extract top hit
IDblast = blastName2ID(blastFile)
# # (2) extract cho, human from gene2refseq
org1ref = extract_from_gene2ref(gene2refseq,taxID1,organism1,columnNum=[2,6,7])
org2ref = extract_from_gene2ref(gene2refseq,taxID2,organism2,columnNum=[2,6,7])
# # (3) replace protein ID with gene ID
IDmap = prIDMap2geneIDMap(IDblast,org1ref,org2ref)
# # (4) sort file by 2nd column
sortFile = IDmap[:-3] + 'sort.txt'
cmd = ('sort -k2,2 -n {input} > {output}').format(input=IDmap,output=sortFile)
subprocess.call(cmd,shell=True)
os.remove(IDblast)
# # (5) now come to the parsing rosetta stone pairs
filename = sortFile
res = open(filename,'r')
inter = 'inter.txt'
interOut = open(inter,'w')
outputFile = filename[:-16] + 'rosetta.txt'
output = open(outputFile,'w')
# combine lines with same 2 columns.
id_pair = res.readline()[:-1].split('\t')
# start from second line
for line in res:
line = line[:-1]
item = line.split('\t')
# should merge
if id_pair[0] == item[0] and id_pair[1] == item[1]:
id_pair[8] = str(min(int(id_pair[8]),int(id_pair[9]),int(item[8]),int(item[9])))
id_pair[9] = str(max(int(id_pair[8]),int(id_pair[9]),int(item[8]),int(item[9])))
else:
interOut.write('\t'.join(id_pair) + '\n')
id_pair = item
interOut.write('\t'.join(id_pair) + '\n')
interOut.close()
res.close()
# now interOut has unique first 2 columns of blast tabular results
res = open(inter,'r')
group = [res.readline()[:-1].split('\t')] # group stores lines with same reference
for line in res:
item = line[:-1].split('\t')
if item[1] == group[-1][1]:
group.append(item)
else:
# whether to do rosetta pair detection
if len(group) > 1:
rosetta = rosetta_pair(group)
# output to file
if rosetta != []:
for pair in rosetta:
output.write('\t'.join(pair) + '\n')
# there is no rosetta pairs
group = [item]
if len(group) > 1:
rosetta = rosetta_pair(group)
# output to file
if rosetta != []:
for pair in rosetta:
output.write('\t'.join(pair) + '\n')
output.close()
res.close()
return outputFile
