def score_gene_pair(self, geneA, geneB, eventsA, eventsB, annotations): (historiesA, costsA) =merge_histories_costs(eventsA) (historiesB, costsB) =merge_histories_costs(eventsB) eventsAB=[] for event in eventsA: genelist = annotations[event.id] if geneA in genelist and geneB in genelist: eventsAB.append(event) if len(genelist) >2: # there are other genes besides A and B self.countsABplus+=1 (self.hists, self.costs) = merge_histories_costs(eventsAB) self.count = len(self.hists) self.countsA =len(historiesA) self.countsB = len(historiesB) self.scoreA = histseg.compute_likelihood(costsA, 1) self.scoreB = histseg.compute_likelihood(costsB, 1) self.likelihood= histseg.compute_likelihood(self.costs, 1)
def main(allevents, annotationfn, totalp, outputfh):
annotations=genemod.read_in_annotations(annotationfn)
sys.stderr.write("Read in all the annotations: %d\n" % (len(annotations)))
geneEvents=genemod.create_gene_events_hash(allevents, annotations) #key: a gene name, value: a list of event indexes
sys.stderr.write("geneEvents: %d\n" % (len(geneEvents)))
if not totalp:
totalp = histseg.get_total_likelihood(allevents)
sys.stderr.write("totalp: %s\n" % (str(totalp)))
for gene in geneEvents.keys():
myevents=[allevents[i] for i in geneEvents[gene]]
sys.stderr.write("Working on %s with %d events\n" % (gene, len(myevents)))
mergedEvent=histseg.merge_events(myevents)
mergedEvent.compute_timing_wmeansd()
mergedEvent.likelihood = histseg.compute_likelihood(mergedEvent.costs, totalp)
mergedEvent.numhists=len(mergedEvent.histories)
self = mergedEvent
mystr=("%s\t%f\t%s\t%f\t%s\t%s\t%d\n" % (gene, self.ordermean, str(self.ordersd), self.prevalmean, str(self.prevalsd), str(self.likelihood), self.numhists))
outputfh.write(mystr)
sys.stderr.write(mystr)
def likelihood_score_gene_pairs(allevent, annotations, tabixfn):
geneEvents=create_gene_events_hash(allevents, annotations) # key: a gene name, value: a list of event indexes for the events with geneX
sys.stderr.write("Hashed in all the annotations: %d\n" % (len(geneEvents)))
eventi=0
myGenepairs=[]
mygeneunpairs=[]
pairIDs=[]
allhistoryids=[]
allcosts=[]
while eventi < len(allevents):
sys.stderr.write("working on event %d\n" % (eventi))
myevent=allevents[eventi]
for i in xrange(len(myevent.histories)):
hid=myevent.histories[i]
if hid not in allhistoryids:
allhistoryids.append(hid)
allcosts.append(myevent.costs[i])
genes=annotations[myevent.id]
if genes[0] != "None":
for ia in xrange(len(genes)):
geneA = genes[ia]
for ib in xrange(ia+1, len(genes)):
geneB = genes[ib]
genepairID="%s,%s" % (geneA, geneB)
# sys.stderr.write("working on %s\n" % (genepairID))
if genepairID not in pairIDs:
eventsA=[allevents[i] for i in geneEvents[geneA]]
eventsB=[allevents[i] for i in geneEvents[geneB]]
mypair = GenePair(geneA, geneB, eventsA, eventsB, annotations)
pairIDs.append(genepairID)
myGenepairs.append(mypair)
eventi+=1
totalp = histseg.compute_likelihood(allcosts, 1)
sys.stderr.write("totalp: %s\n" % (str(totalp)))
if tabixfn:
pair_distances = find_distance_between_genes(pairIDs, args.tabix)
for pair in myGenepairs:
if tabixfn: pair.distance = pair_distances[pair.geneA+pair.geneB]
pair.adjust_likelihoods(totalp)
sys.stdout.write(str(pair))