def create_genetree_from_crossdata(crossdata):
"""
(Try to) create the GeneTreeGraph from crossdata dictionairy
@type crossdata: dict
@param crossdata: crossdata <dict data structure>
@rtype: GeneTreeGraph
@return: estimated! GeneTreeGraph constructed from PabcPs with highest bitscore
"""
# fill GeneTreeGraph with nodes
GTG = GeneTreeGraph()
for (geneA,geneB) in crossdata.keys():
if geneA not in GTG.get_nodes():
GTG.add_node(geneA)
if geneB not in GTG.get_nodes():
GTG.add_node(geneB)
# fill GeneTreeGraph with edges
for (geneA,geneB) in crossdata.keys():
keys = crossdata[(geneA,geneB)]['accepted_pacbs'].keys()
if keys:
keys.sort()
keys.reverse()
bestpacbp = crossdata[(geneA,geneB)]['accepted_pacbs'][keys[0]]
# store this edge to the GeneTreeGraph
GTG.add_edge(geneA,geneB,bestpacbp.identityscore)
else:
# no keys at all, meaning no Pacbps between 2 species meaining
# GTG can not be created yet! Set GSG back to empty graph and return
# set GTG back to an empty graph, break out anf return
GTG = GeneTreeGraph()
break
# return the GeneTreeGraph
return GTG
def detect_and_remove_synteny(inwpcbgs,
PCG,
GENE_IDENTIFIER_SET,
verbose=True):
""" """
MIN_OBSERVED_VS_EXPECTED_RATIO = 0.20
observed_organism_subcombis = []
syntenic_subinwpcbgs = []
# detect syntenic genes in MAIN inwpCBGs,
# without taking strongest informants by GTG analyses
syntenic_inwpcbgs = assign_syntenic_inwpcbgs(inwpcbgs)
for syntinwpcbg in syntenic_inwpcbgs:
syntenic_subinwpcbgs.append(syntinwpcbg)
for inwpCBG in inwpcbgs:
# omit inwpCBGs with annotated exons/orfs
if inwpCBG.count_orfs_labeled_as_annotated_exon() >= 2: continue
target = inwpCBG._get_target_organism()
# make a (artificially fully connected) GeneTreeGraph
gtg = GeneTreeGraph()
gtg.add_node(target)
for (pacbpkey, nodeQ, nodeS), pacbporf in inwpCBG.pacbps.iteritems():
orgS = inwpCBG.organism_by_node(nodeS)
if orgS not in GENE_IDENTIFIER_SET: continue
gtg.add_node(orgS)
for (pacbpkey, nodeQ, nodeS), pacbporf in inwpCBG.pacbps.iteritems():
orgQ = inwpCBG.organism_by_node(nodeQ)
orgS = inwpCBG.organism_by_node(nodeS)
if orgS not in GENE_IDENTIFIER_SET: continue
gtg.add_edge(orgQ, orgS, wt=pacbporf.bitscore)
# make artificially missed edges between the informants
for org in inwpCBG.organism_set():
if org not in [orgQ, orgS] and org in GENE_IDENTIFIER_SET:
if gtg.has_edge( orgS, org ) and\
gtg.weights[(orgS, org)] > pacbporf.bitscore:
gtg.set_edge_weight(orgS, org, wt=pacbporf.bitscore)
else:
gtg.add_edge(orgS, org, wt=pacbporf.bitscore)
# omit (nearly) empty genetreegraphs
if gtg.node_count() <= 1: continue
# remove (much) weaker connected nodes as expected from the gtg
while gtg.get_nodes() and MIN_OBSERVED_VS_EXPECTED_RATIO >\
min( [ gtg.get_node_weighted_connectivity_observed_vs_expected(node) for node in gtg.get_nodes() ]):
node = gtg.weakest_connected_node()
gtg.del_node(node)
# check if already tested before; present in observed_organism_subcombis
if gtg.get_ordered_nodes() in observed_organism_subcombis: continue
# store to already tested organism subcombinations
observed_organism_subcombis.append(gtg.get_ordered_nodes())
# create a subPCG of these organisms
subPCG = PacbpCollectionGraph(crossdata={},
blastmatrix=PCG._blastmatrix)
for (pacbpkey, nodeQ, nodeS), pacbporf in PCG.pacbps.iteritems():
(orgQ, orfQid), (orgS, orfSid) = nodeQ, nodeS
if orgQ not in gtg.get_nodes(): continue
if orgS not in gtg.get_nodes(): continue
subPCG.add_node(nodeQ)
subPCG.add_node(nodeS)
subPCG.add_edge(nodeQ, nodeS, wt=pacbporf.bitscore)
subPCG.pacbps[(pacbpkey, nodeQ, nodeS)] = pacbporf
# make inwpCBGs of this subPCG
subinwpcbgs = PCG2inwpCBGS(subPCG)
# check if there are subinwpcbgs
if not subinwpcbgs: continue
########################################################################
#if verbose:
# print "subPCG organism set:", gtg.get_ordered_nodes()
# print_inwpcbgstructure(subinwpcbgs,gtg.get_ordered_nodes())
########################################################################
# create a subInwardsPointingCodingBlockGraph of these organisms
#subinwpCBG = InwardsPointingCodingBlockGraph()
#for (pacbpkey,nodeQ,nodeS), pacbporf in inwpCBG.pacbps.iteritems():
# (orgQ,orfQid),(orgS,orfSid) = nodeQ,nodeS
# if orgQ not in gtg.get_nodes(): continue
# if orgS not in gtg.get_nodes(): continue
# subinwpCBG.add_node(nodeQ)
# subinwpCBG.add_node(nodeS)
# subinwpCBG.add_edge(nodeQ,nodeS,wt=pacbporf.bitscore)
# subinwpCBG.pacbps[(pacbpkey,nodeQ,nodeS)] = pacbporf
# detect syntenic genes in this subinwpcbgs
syntenic_inwpcbgs = assign_syntenic_inwpcbgs(subinwpcbgs)
for syntinwpcbg in syntenic_inwpcbgs:
syntenic_subinwpcbgs.append(syntinwpcbg)
####################################################################
if verbose:
print "SYNTENIC!!", syntinwpcbg, syntinwpcbg.get_ordered_nodes(
)
for subCBG in subinwpcbgs:
print "syntenic in:", subCBG, subCBG.get_ordered_nodes()
####################################################################
if not syntenic_subinwpcbgs:
return False
# cleanup all inwpCBGs from the syntenic subInwpCBGs
syntenic_pacbpkeys = []
for syntinwpcbg in syntenic_subinwpcbgs:
node_set = syntinwpcbg.node_set()
for inwpCBG in inwpcbgs:
if not node_set.difference(inwpCBG.node_set()):
for pacbpkey in inwpCBG.pacbps.keys():
if pacbpkey not in syntenic_pacbpkeys:
syntenic_pacbpkeys.append(pacbpkey)
# place all syntenic_pacbpkeys and PacbPORFs in the syntenicPCG
# and, at the same time, remove from the main PCG
syntenicPCG = PacbpCollectionGraph(crossdata={},
blastmatrix=PCG._blastmatrix)
for key in syntenic_pacbpkeys:
(pacbpkey, nodeQ, nodeS) = key
pacbporf = PCG.pacbps[key]
# add to syntenicPCG
syntenicPCG.add_node(nodeQ)
syntenicPCG.add_node(nodeS)
syntenicPCG.add_edge(nodeQ, nodeS, wt=pacbporf.bitscore)
syntenicPCG.pacbps[(pacbpkey, nodeQ, nodeS)] = pacbporf
# remove from main PCG
_delete_pacbp(PCG, key)
# return syntenicPCG
return syntenicPCG
def detect_and_remove_synteny(inwpcbgs,PCG,GENE_IDENTIFIER_SET,verbose=True):
""" """
MIN_OBSERVED_VS_EXPECTED_RATIO = 0.20
observed_organism_subcombis = []
syntenic_subinwpcbgs = []
# detect syntenic genes in MAIN inwpCBGs,
# without taking strongest informants by GTG analyses
syntenic_inwpcbgs = assign_syntenic_inwpcbgs(inwpcbgs)
for syntinwpcbg in syntenic_inwpcbgs:
syntenic_subinwpcbgs.append(syntinwpcbg)
for inwpCBG in inwpcbgs:
# omit inwpCBGs with annotated exons/orfs
if inwpCBG.count_orfs_labeled_as_annotated_exon() >= 2: continue
target = inwpCBG._get_target_organism()
# make a (artificially fully connected) GeneTreeGraph
gtg = GeneTreeGraph()
gtg.add_node(target)
for (pacbpkey,nodeQ,nodeS),pacbporf in inwpCBG.pacbps.iteritems():
orgS = inwpCBG.organism_by_node(nodeS)
if orgS not in GENE_IDENTIFIER_SET: continue
gtg.add_node(orgS)
for (pacbpkey,nodeQ,nodeS),pacbporf in inwpCBG.pacbps.iteritems():
orgQ = inwpCBG.organism_by_node(nodeQ)
orgS = inwpCBG.organism_by_node(nodeS)
if orgS not in GENE_IDENTIFIER_SET: continue
gtg.add_edge( orgQ, orgS, wt = pacbporf.bitscore )
# make artificially missed edges between the informants
for org in inwpCBG.organism_set():
if org not in [orgQ,orgS] and org in GENE_IDENTIFIER_SET:
if gtg.has_edge( orgS, org ) and\
gtg.weights[(orgS, org)] > pacbporf.bitscore:
gtg.set_edge_weight(orgS,org,wt = pacbporf.bitscore)
else:
gtg.add_edge( orgS, org, wt = pacbporf.bitscore )
# omit (nearly) empty genetreegraphs
if gtg.node_count() <= 1: continue
# remove (much) weaker connected nodes as expected from the gtg
while gtg.get_nodes() and MIN_OBSERVED_VS_EXPECTED_RATIO >\
min( [ gtg.get_node_weighted_connectivity_observed_vs_expected(node) for node in gtg.get_nodes() ]):
node = gtg.weakest_connected_node()
gtg.del_node(node)
# check if already tested before; present in observed_organism_subcombis
if gtg.get_ordered_nodes() in observed_organism_subcombis: continue
# store to already tested organism subcombinations
observed_organism_subcombis.append( gtg.get_ordered_nodes() )
# create a subPCG of these organisms
subPCG = PacbpCollectionGraph(crossdata={},
blastmatrix=PCG._blastmatrix)
for (pacbpkey,nodeQ,nodeS), pacbporf in PCG.pacbps.iteritems():
(orgQ,orfQid),(orgS,orfSid) = nodeQ,nodeS
if orgQ not in gtg.get_nodes(): continue
if orgS not in gtg.get_nodes(): continue
subPCG.add_node(nodeQ)
subPCG.add_node(nodeS)
subPCG.add_edge(nodeQ,nodeS,wt=pacbporf.bitscore)
subPCG.pacbps[(pacbpkey,nodeQ,nodeS)] = pacbporf
# make inwpCBGs of this subPCG
subinwpcbgs = PCG2inwpCBGS(subPCG)
# check if there are subinwpcbgs
if not subinwpcbgs: continue
########################################################################
#if verbose:
# print "subPCG organism set:", gtg.get_ordered_nodes()
# print_inwpcbgstructure(subinwpcbgs,gtg.get_ordered_nodes())
########################################################################
# create a subInwardsPointingCodingBlockGraph of these organisms
#subinwpCBG = InwardsPointingCodingBlockGraph()
#for (pacbpkey,nodeQ,nodeS), pacbporf in inwpCBG.pacbps.iteritems():
# (orgQ,orfQid),(orgS,orfSid) = nodeQ,nodeS
# if orgQ not in gtg.get_nodes(): continue
# if orgS not in gtg.get_nodes(): continue
# subinwpCBG.add_node(nodeQ)
# subinwpCBG.add_node(nodeS)
# subinwpCBG.add_edge(nodeQ,nodeS,wt=pacbporf.bitscore)
# subinwpCBG.pacbps[(pacbpkey,nodeQ,nodeS)] = pacbporf
# detect syntenic genes in this subinwpcbgs
syntenic_inwpcbgs = assign_syntenic_inwpcbgs(subinwpcbgs)
for syntinwpcbg in syntenic_inwpcbgs:
syntenic_subinwpcbgs.append(syntinwpcbg)
####################################################################
if verbose:
print "SYNTENIC!!", syntinwpcbg, syntinwpcbg.get_ordered_nodes()
for subCBG in subinwpcbgs:
print "syntenic in:", subCBG, subCBG.get_ordered_nodes()
####################################################################
if not syntenic_subinwpcbgs:
return False
# cleanup all inwpCBGs from the syntenic subInwpCBGs
syntenic_pacbpkeys = []
for syntinwpcbg in syntenic_subinwpcbgs:
node_set = syntinwpcbg.node_set()
for inwpCBG in inwpcbgs:
if not node_set.difference(inwpCBG.node_set()):
for pacbpkey in inwpCBG.pacbps.keys():
if pacbpkey not in syntenic_pacbpkeys:
syntenic_pacbpkeys.append(pacbpkey)
# place all syntenic_pacbpkeys and PacbPORFs in the syntenicPCG
# and, at the same time, remove from the main PCG
syntenicPCG = PacbpCollectionGraph(crossdata={},blastmatrix=PCG._blastmatrix)
for key in syntenic_pacbpkeys:
(pacbpkey,nodeQ,nodeS) = key
pacbporf = PCG.pacbps[key]
# add to syntenicPCG
syntenicPCG.add_node(nodeQ)
syntenicPCG.add_node(nodeS)
syntenicPCG.add_edge(nodeQ,nodeS,wt=pacbporf.bitscore)
syntenicPCG.pacbps[(pacbpkey,nodeQ,nodeS)] = pacbporf
# remove from main PCG
_delete_pacbp(PCG,key)
# return syntenicPCG
return syntenicPCG