def __init__(self, biosettings, id=15):
bioloader.BioLoader.__init__(self, biosettings, id)
self.biosettings.PurgeGroupData(id)
biosettings.LoadAliases()
self.kb = bioloader.KnowledgeBase(id, "PharmGKB",
"Pharmacogenomics Knowledge Base")
self.pathwayGroups = bioloader.Pathway(self.kb.groupTypeID, id + 1,
"Pathways", "PharmGKB Pathways")
self.kb.AddPathway(self.pathwayGroups, "PharmGKB Pathway")
self.pathways = []
#self.diseaseGroups = bioloader.KnowledgeBase(id+1, "PharmGKB-Diseases", "Pharmacogenomics Knowledge Base (Disease)")
self.diseaseGroups = bioloader.Pathway(
self.kb.groupTypeID, id + 2, "Diseases",
"PharmGKB Disease Associations")
self.kb.AddPathway(self.diseaseGroups, "PharmGKB Disease")
#self.drugGroups = bioloader.KnowledgeBase(id+2, "PharmGKB-Drugs", "Pharmacogenomics Knowledge Base (Drug)")
self.drugGroups = bioloader.Pathway(self.kb.groupTypeID, id + 3,
"Drugs",
"PharmGKB Drug Associations")
self.kb.AddPathway(self.drugGroups, "PharmGKB Drug")
self.gkbentities = dict()
self.gkbentities["Gene"] = dict()
self.gkbentities["Drug"] = dict()
self.gkbentities["Disease"] = dict()
def CommitToDB(self, kb, parentGroup):
if kb.groupTypeID not in self.alreadyCommitted:
self.alreadyCommitted.add(kb.groupTypeID)
curPathway = bioloader.Pathway(kb.groupTypeID, self.GetID(),
self.acc,
"%s (%s)" % (self.name, self.type))
parentGroup.AddAssociation(curPathway.groupID, "")
#kb.AddPathway(curPathway)
#curPathway.AssociatePathways(group_id, self.id, self.type)
if "Drug" in self.associations:
for item in self.associations["Drug"]:
#print item
#print >> sys.stderr, kb.groupTypeID, "-> ", item.name
curPathway.AddAssociation(item.GetID(), "Drug")
item.CommitToDB(kb, curPathway)
if "Disease" in self.associations:
for item in self.associations["Disease"]:
#print >> sys.stderr, kb.groupTypeID, "-> ", item.name
curPathway.AddAssociation(item.GetID(), "Disease")
item.CommitToDB(kb, curPathway)
if "Gene" in self.associations:
#print "Saving genes to the database, %s in total" % (len(self.associations["Gene"]))
for item in self.associations["Gene"]:
#print >> sys.stderr, curPathway.groupID, " + ", item.name
item.AddToPathway(curPathway)
curPathway.Commit(self.biosettings)
def LoadPathways(self, filename):
curPathway = None
for line in open(filename):
words = line.split(":")
if len(words) > 1:
newID = self.biosettings.NextID()
if curPathway:
self.pathways.append(curPathway)
curPathway = bioloader.Pathway(self.groupID, newID,
words[0].strip(),
words[1].strip())
self.pathwayGroups.AddAssociation(curPathway.groupID,
"Pathway")
#self.kb.AssociatePathways(self.groupID, newID, "Member")
else:
#pass
words = line.split()
if len(words) > 1 and words[0] == "Gene":
geneIDs = self.biosettings.regions.AliasToGeneID(
[words[2].strip()])
if len(geneIDs) == 1:
curPathway.AddGene(geneIDs[0])
if curPathway:
self.pathways.append(curPathway)
os.system("mv %s archives" % (filename))
def LoadAssociation(self, line):
cols = line.split("\t")
success = False
if len(cols[0].split(":")) > 1:
idA = cols[0].split(":") #type:id
idB = cols[1].split(":")
geneA = None
geneB = None
acceptableTypes = ["uniprotkb"]
taxA = cols[9].split(":")[-1].strip()
taxB = cols[10].split(":")[-1].strip()
idLookup = BuildLookup(cols[13])
if taxA == taxB and taxA == "9606(H**o sapiens)":
#we can also attempt to id a gene using columns 4 & 5
if idA[0] in acceptableTypes:
ids = self.biosettings.regions.AliasToGeneID([idA[1]])
if len(ids) == 1:
geneA = list(ids)[0]
elif len(ids) == 0:
self.missingUni = idA[1]
else:
self.ambiguousUni = idA[1]
if idB[0] in acceptableTypes:
ids = self.biosettings.regions.AliasToGeneID([idB[1]])
if len(ids) == 1:
geneB = list(ids)[0]
elif len(ids) == 0:
self.missingUni = idB[1]
else:
self.ambiguousUni = idB[1]
if geneA != None and geneB != None and geneA != geneB:
groupID = self.biosettings.NextID()
if idLookup["mint"] in self.observedIDs:
print "!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!"
print thisdoesnt.exit
self.observedIDs.add(idLookup["mint"])
pathway = bioloader.Pathway(self.groupID, groupID,
idLookup["mint"], "")
pathway.AddGene(geneA)
pathway.AddGene(geneB)
pathway.Commit(self.biosettings)
self.biosettings.RelatePathways(self.groupID, groupID,
"mint", "")
success = True
#else:
# print "Unknown tax ID: ", taxA, taxB
return success
def Commit(self, groupTypeID, biosettings):
#First, we'll add ourselves to the groups table
pathway = bioloader.Pathway(groupTypeID, self.groupID, self.term_id,
self.description)
#Next, add our associations to the association table
for association in self.associations:
pathway.AddGene(association[0])
#finally, we can add the various groups. If we have no parents, we can set ourselves as our own parent (since we don't have any root)
parentCount = len(self.parents)
for parent in self.parents:
biosettings.RelatePathways(parent, self.groupID,
self.parents[parent][0],
self.parents[parent][1])
if parentCount == 0:
biosettings.RelatePathways(groupTypeID, self.groupID, "", "")
pathway.Commit(biosettings)
def LoadAssociation(self, line):
cols = line.split("\t")
words = cols[0].split("|")
dipA = words[0].strip()
uniA = words[len(words)-1][10:]
words = cols[1].split("|")
dipB = words[0].strip()
uniB = words[len(words)-1][10:]
taxA = cols[9][6:]
taxB = cols[10][6:]
comment = cols[6]
name = cols[13]
success = False
if taxA == taxB:
idsA = self.biosettings.regions.AliasToGeneID([uniA])
idsB = self.biosettings.regions.AliasToGeneID([uniB])
if len(idsA) == 1 and len(idsB) == 1:
geneA = list(idsA)[0]
geneB = list(idsB)[0]
if geneA != geneB:
groupID = self.biosettings.NextID()
pathway = bioloader.Pathway(self.groupID, groupID, name, comment)
pathway.AddGene(geneA)
pathway.AddGene(geneB)
pathway.Commit(self.biosettings)
self.biosettings.RelatePathways(self.groupID, groupID, "PFAM", "")
success = True
else:
if len(idsA) == 0:
self.missingUni.add(uniA)
else:
self.ambiguousUni.add(uniA)
if len(idsB) == 0:
self.missingUni.add(uniB)
else:
self.ambiguousUni.add(uniB)
else:
print >> sys.stdout, "Misaligned Taxonomy (%s), %s %s : %s" % (name, taxA, taxB, comment)
return success
def LoadFile(self, groupID, filename):
#print filename
elements = filename.split("_")
name = "_".join(elements[:2])
elements = name.split("/")
if len(elements) > 1:
name = elements[1]
#print "Pathway Name: ", name
pathway = bioloader.Pathway(self.groupID, groupID, name, filename)
lineCount = 0
geneCount = 0
genesMissed = 0
for line in open(filename):
words = line.strip().split()
if lineCount > 0:
geneID = 0
geneName = words[2].strip()
entrezID = words[3].strip()
geneIDs = self.biosettings.regions.AliasToGeneID(
[entrezID, geneName])
if len(geneIDs) > 0:
if len(geneIDs) == 1:
pathway.AddGene(geneIDs.pop())
geneCount += 1
else:
pass
#print "-----(%s,%s)\tUnable to disambiguate gene ID (%s distinct ids) " % (geneName, entrezID, len(geneIDs))
else:
pass
#print "-----(%s,%s)\tUnable to recognize gene" % (entrezID, geneName)
genesMissed += 1
lineCount += 1
#print "Lines in File: %s\nGenes Identified: %s\nGenes Unable to be Identified: %s" % (lineCount, geneCount, genesMissed)
pathway.Commit(self.biosettings)
self.biosettings.RelatePathways(self.groupID, groupID, "", "")
def LoadPathway(self, groupID, remotePathway):
definition = remotePathway['definition'][0]
entryID = remotePathway['entry_id'][0]
#print remotePathway
genes = set() # geneIDs identified with pathway
failedInsertions = 0
geneCount = 0
#print "\n\nPathway(%s): %s - %s" % (groupID, entryID, definition)
#because this could timeout, let's make sure it doesn't just faile
doContinue = False
while not doContinue:
#try
remoteGeneList = self.srv.get_genes_by_pathway(pathway_id=entryID)
doContinue = True
for geneHSA in remoteGeneList:
gene = geneHSA[4:]
geneIDs = self.biosettings.regions.AliasToGeneID([gene])
if len(geneIDs) == 0:
#print "-----(%s, %s)\tUnable to recognize gene" % (gene, geneHSA)
failedInsertions += 1
elif len(geneIDs) != 1:
failedInsertions += 1
#print "-----(%s,%s)\tUnable to disambiguate gene ID (%s distinct ids) " % (gene, geneHSA, len(geneIDs))
else:
genes.add(geneIDs[0])
if len(genes) > 0:
pathway = bioloader.Pathway(self.groupID, groupID, entryID[5:],
definition)
for geneID in genes:
pathway.AddGene(geneID)
geneCount += 1
pathway.Commit(self.biosettings)
self.biosettings.RelatePathways(self.groupID, groupID, "", "")
def LoadAssociation(self, line):
#header line
if line[0] == "#":
return False
cols = line.split("\t")
success = False
geneA = int(cols[1])
geneB = int(cols[2])
name = "biogrid:%s" % (cols[0])
cantContinue = False
taxA = cols[15].strip()
taxB = cols[16].strip()
if taxA == taxB and taxA == "9606":
if geneA not in self.biosettings.regions.genes:
self.missing.add(geneA)
cantContinue = True
if geneB not in self.biosettings.regions.genes:
self.missing.add(geneB)
cantContinue = True
if not cantContinue:
groupID = self.biosettings.NextID()
pathway = bioloader.Pathway(self.groupID, groupID, name, "")
pathway.AddGene(geneA)
pathway.AddGene(geneB)
pathway.Commit(self.biosettings)
self.biosettings.RelatePathways(self.groupID, groupID,
cols[12].strip(), "")
success = True
#else:
# print "Unknown tax ID: ", taxA, taxB
else:
pass
#print "Unable to find one or more genes (%s, %s) in %s total genes" % (geneA, geneB, len(self.biosettings.regions.genes))
return success