def main(BgShelve, BgFile, BgGeneIDCol, BgTermCol, BgStartRow1, FgFile,
FgGeneIDCol, FgTermCol, FgStartRow1, recomputeBg):
global termSeperator
if not recomputeBg:
saved = shelve.open(BgShelve)
if not saved.has_key("BG"):
recomputeBg = True
else:
BGData = saved["BG"]
if len(BGData) < 1 or not BGData.has_key("__$$$NGENES"):
recomputeBg = True
saved.close()
if recomputeBg:
print >> stderr, "recal BG"
BGData = dict()
#=shelve.open(BgShelve);
calTerms(BGData, BgFile, BgGeneIDCol, BgTermCol, BgStartRow1)
saved = shelve.open(BgShelve)
saved["BG"] = BGData
saved.close()
FGData = dict()
calTerms(FGData, FgFile, FgGeneIDCol, FgTermCol, FgStartRow1)
#now calculate GO enrichment and FDR
FGTerms = FGData.keys()
nTermsBG = BGData["__$$$NTERMS"]
nGenesBG = BGData["__$$$NGENES"]
nTermsFG = FGData["__$$$NTERMS"]
nGenesFG = FGData["__$$$NGENES"]
pvalueTermMap = dict()
for term in FGTerms:
if "__$$$" in term:
continue
if not BGData.has_key(term):
print >> stderr, "Error: term", term, "not found in background"
BGTermEntry = BGData[term]
FGTermEntry = FGData[term]
pop = nGenesBG
popt = len(BGTermEntry)
sam = nGenesFG
samt = len(FGTermEntry)
pvalue = pvalue_enrichment(pop, popt, sam, samt)
if not pvalueTermMap.has_key(pvalue):
pvalueTermMap[pvalue] = []
resultVector = [term, pop, popt, sam, samt, FGTermEntry]
pvalueTermMap[pvalue].append(resultVector)
print >> stderr, resultVector, ",pvalue=", pvalue
#now sort pvalue and calculate FDR
pvalues = sorted(pvalueTermMap.keys())
nInc = 0
print >> stdout, "FDR", "\t",
print >> stdout, "p-value", "\t",
print >> stdout, "term", "\t",
print >> stdout, "popt", "\t",
print >> stdout, "pop", "\t",
print >> stdout, "popt/pop", "\t",
print >> stdout, "samt", "\t",
print >> stdout, "sam", "\t",
print >> stdout, "samt/sam", "\t",
print >> stdout, "genes", "\t"
for pvalue in pvalues:
termsWithThisPvalue = pvalueTermMap[pvalue]
nTermsWithThisPvalue = len(termsWithThisPvalue)
nInc += nTermsWithThisPvalue
FDR = (pvalue * nTermsFG) / nInc
#now print;
for termEntry in termsWithThisPvalue:
term, pop, popt, sam, samt, FGTermEntry = termEntry
print >> stdout, FDR, "\t",
print >> stdout, pvalue, "\t",
print >> stdout, term, "\t",
print >> stdout, popt, "\t",
print >> stdout, pop, "\t",
print >> stdout, float(popt) / pop, "\t",
print >> stdout, samt, "\t",
print >> stdout, sam, "\t",
print >> stdout, float(samt) / sam, "\t",
print >> stdout, ",".join(FGTermEntry), "\t"
def main(BgShelve, BgFile, BgGeneIDCol, BgTermCol, BgStartRow1, FgFile, FgGeneIDCol, FgTermCol,FgStartRow1,recomputeBg):
global termSeperator;
if not recomputeBg:
saved=shelve.open(BgShelve);
if not saved.has_key("BG"):
recomputeBg=True;
else:
BGData=saved["BG"];
if len(BGData)<1 or not BGData.has_key("__$$$NGENES"):
recomputeBg=True;
saved.close();
if recomputeBg:
print >> stderr, "recal BG";
BGData=dict();#=shelve.open(BgShelve);
calTerms(BGData,BgFile,BgGeneIDCol,BgTermCol,BgStartRow1);
saved=shelve.open(BgShelve);
saved["BG"]=BGData;
saved.close();
FGData=dict();
calTerms(FGData,FgFile,FgGeneIDCol,FgTermCol,FgStartRow1);
#now calculate GO enrichment and FDR
FGTerms=FGData.keys();
nTermsBG=BGData["__$$$NTERMS"];
nGenesBG=BGData["__$$$NGENES"];
nTermsFG=FGData["__$$$NTERMS"];
nGenesFG=FGData["__$$$NGENES"];
pvalueTermMap=dict();
for term in FGTerms:
if "__$$$" in term:
continue;
if not BGData.has_key(term):
print >> stderr, "Error: term",term,"not found in background";
BGTermEntry=BGData[term];
FGTermEntry=FGData[term];
pop=nGenesBG;
popt=len(BGTermEntry);
sam=nGenesFG;
samt=len(FGTermEntry);
pvalue=pvalue_enrichment(pop,popt,sam,samt);
if not pvalueTermMap.has_key(pvalue):
pvalueTermMap[pvalue]=[];
resultVector=[term,pop,popt,sam,samt,FGTermEntry]
pvalueTermMap[pvalue].append(resultVector);
print >> stderr, resultVector,",pvalue=",pvalue;
#now sort pvalue and calculate FDR
pvalues=sorted(pvalueTermMap.keys());
nInc=0;
print >> stdout, "FDR","\t",
print >> stdout, "p-value","\t",
print >> stdout, "term","\t",
print >> stdout, "popt","\t",
print >> stdout, "pop", "\t",
print >> stdout, "popt/pop","\t",
print >> stdout, "samt","\t",
print >> stdout, "sam","\t",
print >> stdout, "samt/sam", "\t",
print >> stdout, "genes","\t";
for pvalue in pvalues:
termsWithThisPvalue=pvalueTermMap[pvalue];
nTermsWithThisPvalue=len(termsWithThisPvalue);
nInc+=nTermsWithThisPvalue;
FDR=(pvalue*nTermsFG)/nInc;
#now print;
for termEntry in termsWithThisPvalue:
term,pop,popt,sam,samt,FGTermEntry=termEntry;
print >> stdout, FDR,"\t",
print >> stdout, pvalue,"\t",
print >> stdout, term,"\t",
print >> stdout, popt,"\t",
print >> stdout, pop, "\t",
print >> stdout, float(popt)/pop,"\t",
print >> stdout, samt,"\t",
print >> stdout, sam,"\t",
print >> stdout, float(samt)/sam, "\t",
print >> stdout, ",".join(FGTermEntry),"\t";
#!/usr/bin/python from hypergeom import pvalue_enrichment; for i in range(1,1000): r=pvalue_enrichment(19507150 , 11324 , 22457 , 17); print r;