def file_parser():
"""from bob"s proteinGroups.txt take: Majority protein IDs Peptide counts (razor+unique) ['LFQ intensity KO1', 'LFQ intensity KO2', 'LFQ intensity KO3', 'LFQ intensity WT1', 'LFQ intensity WT2', 'LFQ intensity WT3']
and write them to a new file. do not select contaminants or reverse peptides"""
from root.ed.tools import file_importer, file_outporter
print "this is file parser"
inpF = file_importer("bob/24h_proteingroups.csv")
outF = file_outporter("bob/processed/24h_bobdata.csv")
for inpLine in inpF:
inpP = inpLine.split("\r")
cN = 0
print len(inpP)
for inpI in inpP:
inpItems = inpI.split("\t")
if inpItems[100] == "+" or inpItems[101] == "+": continue # get rid of contaminants and reverse proteins
outF.write(str(cN) + "," + inpItems[1] + "," + inpItems[6] + "," + inpItems[3] + "," + inpItems[86] + "," + inpItems[87] + "," + inpItems[88] + "," + inpItems[89] + "," + inpItems[90] + "," + inpItems[91] + "\n")
# print inpItems [1],"+++", inpItems [3],"+++", inpItems [6],"+++", inpItems[86:92]
cN += 1
# if cN == 40: break
break
inpF.close()
outF.close()
print cN
def entry_parser():
"""remove duplicate protein name and total peptide count cell entries from bob's dataset"""
from root.ed.tools import file_importer, file_outporter
print "this is entry parser"
relPath = "bob/processed/24h_bobdata.csv"
outPath = "bob/processed/24h_bobdata_ed.csv"
inpF = file_importer(relPath)
outF = file_outporter(outPath)
cN = 0
hitN = 0
for inpLine in inpF:
cN += 1
inpLine = inpLine.strip()
inpItem = inpLine.split(",")
geneL = inpItem[1].split(";")
lenS = len(geneL[0])
curGene = geneL[0]
for geneI in geneL:
if len(geneI) < lenS:
lenS = len(geneI)
curGene = geneI
if "__" in curGene: continue
protL = inpItem[3].split(";")
curProt = protL[geneL.index(curGene)]
if curGene[-2] == "-":
curGene = curGene[:-2]
if curGene[-3] == "-":
curGene = curGene[:-3]
outF.write(inpItem[0] + "," + curGene + "," + inpItem[2] + "," + curProt + "," + inpItem[4] + "," + inpItem[5] + "," + inpItem[6] + "," + inpItem[7] + "," + inpItem[8] + "," + inpItem[9] + "\n")
hitN += 1
print cN, hitN
inpF.close()
outF.close()
def stat_parser():
"""take protein names with a significant p value and out them to a result file"""
from root.ed.tools import file_importer, file_outporter
from math import log
print "this is stat parser"
relPath = "bob/processed/24h_bobdata_ed2.csv"
outPathUp = "bob/processed/24h_bobprots_up_full.csv"
outPathDown = "bob/processed/24h_bobprots_down_full.csv"
inpF = file_importer(relPath)
outFUp = file_outporter(outPathUp)
outFDown = file_outporter(outPathDown)
skipFlag = True
for inpLine in inpF:
if skipFlag:
skipFlag = False
outFDown.write("ID,Uniprot ID,Gene name,unique peptides (unique+razor),KO1,KO2,KO3,WT1,WT2,WT3,enrichment,P value\n")
outFUp.write("ID,Uniprot ID,Gene name,unique peptides (unique+razor),KO1,KO2,KO3,WT1,WT2,WT3,enrichment,P value\n")
continue
inpLine = inpLine.split("\" \"")
curLine = []
for inpI in inpLine:
curLine.append(inpI.strip("\"\n"))
try:
curLine[-1] = float(curLine[-1])
except ValueError:
curLine[-1] = 1
if curLine[-1] < 0.05 and int(curLine[3]) > 1: # check if protein has at least 2 unique peptides and has a significant p value
curLine[4:10] = [int(x) for x in curLine[4:10]]
enrScore = log((sum(curLine[4:7]) / 3.0)/(sum(curLine[7:10]) / 3.0),2) # calculate log2 enrichment score
# print int(sum(curLine[4:7]) / 3.0), int(sum(curLine[7:10]) / 3.0)
if sum(curLine[4:7]) / 3.0 > sum(curLine[7:10]) / 3.0: # if the mean of the KO intensities is higher than the wt
for outI in curLine:
outFDown.write(str(outI).strip(" "))
if outI is not curLine[-1]:
outFDown.write(",")
if outI is curLine[-2]:
outFDown.write(str(enrScore)+ ",")
else:
outFDown.write("\n")
# outFDown.write(curLine[1] + "," + curLine[2] + "\n")
else:
# outFUp.write(curLine[1] + "," + curLine[2] + "\n")
for outI in curLine:
outFUp.write(str(outI).strip(" "))
if outI is not curLine[-1]:
outFUp.write(",")
if outI is curLine[-2]:
outFUp.write(str(enrScore)+ ",")
else:
outFUp.write("\n")
inpF.close()
outFUp.close()
outFDown.close()
print "stat_parser completed"
def lfq_parser():
"""remove 0 values from lfq measurements and replace them with a random number between 1 and 100
This is needed for ttseting later in R, as each measurement there has to have some sort of noise in it"""
from root.ed.tools import file_importer, file_outporter
from random import randint
print "this is lfq parser"
relPath = "bob/processed/24h_bobdata_ed.csv"
outPath = "bob/processed/24h_bobdata_no0_ed.csv"
inpF = file_importer(relPath)
outF = file_outporter(outPath)
headerFlag = True
for inpLine in inpF:
if headerFlag:
headerFlag = False
outF.write(inpLine)
continue
inpLine = inpLine.strip()
inpItems = inpLine.split(",")
try:
int(inpItems[4]) # get rid of wonky erroneous lines introduced by excel
except ValueError:
print "bad line found here, ignored: ", inpItems
continue
for inpI in inpItems[0:4]: # copy over gene name and such to new file
outF.write(inpI)
outF.write(",")
commaCount = 0
for inpJ in inpItems[4:]: # copy over lfq values while replacing 0-s with random values
commaCount += 1
if int(inpJ) == 0:
randNum = randint(1,100)
outF.write(str(randNum))
else:
outF.write(inpJ)
if commaCount < 6:
outF.write(",")
outF.write("\n")
inpF.close()
outF.close()
def main():
from root.ed.tools import file_importer, uniprot_dicter, iso_e
from collections import defaultdict
print "this is evidence parser"
inpEvF = file_importer("bob/evidence.txt")
seqD = uniprot_dicter()
resD = defaultdict(list)
for keyS, valueS in seqD.items(): # create dict with keys as all mouse genes and values as a list, with the protein sequence as first item in the list
resD[keyS].append(valueS.strip())
geneL = []
goodC = 0
badC = 0
for inpLine in inpEvF:
inpList = inpLine.split("\t")
inpL = inpList[14].split(";") # this is the leading protein name column. It holds uniprot IDs separated by ;
try:
inpFrac = int(inpList[20]) # this is the "fraction" column. it holds the isoelectric fractionation number
except ValueError:
print inpList[20]
continue
for inpI in inpL:
if inpI[-2] == "-":
inpI = inpI[:-2]
if inpI not in geneL: # make a list of gene names
geneL.append(inpI)
if inpI[:3] == "REV" or inpI[:3] == "CON": continue # there are some weird entries. CON is contaminants methinks.
try:
if inpFrac - 2 < iso_e(resD[inpI][0]) < inpFrac + 2 : # check if the isoelectric point of the protein to which this peptide belongs to matches the fraction it was retrieved from
goodC += 1
else:
badC += 1
"""resD[inpI].append()
resD[inpI].append(iso_e(resD[inpI][0])) # add isoelectric point to the genes found"""
except IndexError:
print "%s has no sequence" % inpI
print str(len(geneL)) + " genes found in evidence.txt"
print str(goodC) + " matching, and " + str(badC) + " unmatching peptides found"
inpEvF.close()
def intact_parser():
"""open ptpn22.txt and extract prey protein uniprot accessions.
Convert those to refseq protein accessions.
Return them as a list."""
from root.ed.tools import prot_id_converter, file_importer
relPath = "ptpn22_ppi_data/ptpn22.txt"
inpF = file_importer(relPath, "r")
headerFlag = True
preyL = []
for inpLine in inpF:
if headerFlag:
headerFlag = False
continue
inpList = inpLine.split("\t")
inpItem = inpList[1].split(":")[-1]
if inpItem not in preyL:
preyL.append(inpItem)
inpItem = inpList[0].split(":")[-1]
if inpItem not in preyL:
preyL.append(inpItem)
inpF.close()
idList = prot_id_converter(preyL, "", outDB="refseqproteingi") # convert uniprot ID to refseq accessions
return idList