def harvest_elegiable_tss_sites(self,max_aa_distance=ALIGNED_TSS_MAX_AA_DISTANCE,
tss_min_pssm_score=TSS_MIN_PSSM_SCORE,
skip_nonelegiable_sites=True):
"""
"""
# update minimal pssm score to stg collection object
stg = TranslationalStartSiteCollectionGraph()
stg.MIN_PSSM_SCORE = tss_min_pssm_score
stg.ALIGNED_SITE_AA_OFFSET = max_aa_distance
# First, proces each individual organism.
for org in self.organism_set():
# take the first (and only) orf of this organism
theorf = self.get_orfs_of_graph(organism=org)[0]
# ready if there are no potential tss loci (no ATG sequence)
if not theorf.has_start(): continue
# scan for tss loci
theorf.scan_orf_for_pssm_tss(min_pssm_score=tss_min_pssm_score)
if skip_nonelegiable_sites:
# get the considered TSS range
(min_aa_pos, min_nt_pos) = self.minimal_eligable_tss_position(org)
(max_aa_pos, max_nt_pos) = self.maximal_eligable_tss_position(org)
else:
(min_aa_pos, min_nt_pos) = None, None
(max_aa_pos, max_nt_pos) = None, None
for tss in theorf._tss_sites:
# check if we can ignore this site
if min_nt_pos and tss.pos < min_nt_pos: continue
if max_nt_pos and tss.pos > max_nt_pos: continue
# an accepted site; add to TSS Collection Graph
startpos = tss.pos / 3
tssNode = ( org, theorf.id, startpos, tss.pos )
stg.add_node_and_object(tssNode,tss)
# Second, evaluate all cross combinations
for ( (a,b,c,d),(g1,o1),(g2,o2) ), pacbporf in self.pacbps.iteritems():
# only proces this combination if both organisms have splice sites!
if g1 not in stg.organism_set(): continue
if g2 not in stg.organism_set(): continue
# now loop over all TSS in Query and Sbjct
# and align them in a graph; an edge is added if 2 sites
# are less then ``max_aa_distance`` apart from each other
for tssQ in stg.get_organism_objects(g1):
# the node that represents this site
startQpos = tssQ.pos / 3
startQnode = ( g1, o1, startQpos, tssQ.pos )
for tssS in stg.get_organism_objects(g2):
# the node that represents this site
startSpos = tssS.pos / 3
startSnode = ( g2, o2, startSpos, tssS.pos )
# get distance between (aligned) start-codons
dist = pacbporf.get_distance_aligned_protein_positions(
query=startQpos,sbjct=startSpos)
# continue if distance between start sites is to big
if dist > max_aa_distance: continue
# calculate binary entropies from both positions
startQpositionPos,phaseQ = pacbporf.dnaposition_query(tssQ.pos,forced_return=True)
startSpositionPos,phaseS = pacbporf.dnaposition_sbjct(tssS.pos,forced_return=True)
entropyQ = pacbporf.alignment_entropy(startQpositionPos,method='left')
entropyS = pacbporf.alignment_entropy(startSpositionPos,method='left')
# calculate a weight from distance between startQpos and startSpos
wt = 1.0 / ( 1.0 + float(dist) )
# check if edge already in graph
if stg.has_edge( startQnode, startSnode ):
_wt = stg.weights[( startQnode, startSnode )]
if wt > _wt:
stg.set_edge_weight( startQnode, startSnode, wt=wt )
# and add binary entropy values
stg._edge_binary_entropies[(startQnode, startSnode)] = (entropyQ,entropyS)
stg._edge_binary_entropies[(startSnode, startQnode)] = (entropyS,entropyQ)
else:
stg.add_edge( startQnode, startSnode, wt=wt )
# and add binary entropy values
stg._edge_binary_entropies[(startQnode, startSnode)] = (entropyQ,entropyS)
stg._edge_binary_entropies[(startSnode, startQnode)] = (entropyS,entropyQ)
# Get tcode data for these start codon nodes
# Assuming that this is indeed the start-codon,
# the stretch of ATG untill max(OMSR) will be coding.
# Take the length of this stretch (in nt) as right/3p/upstream window size
omsr = self.overall_minimal_spanning_range()
for (org,orfid,aaPos,dnaPos) in stg.get_nodes():
theorf = self.get_orfs_of_graph(organism=org)[0]
right_window_size = ( max(omsr[(org,orfid)])+1 - aaPos )*3
# confirm that window size is not < 0; this is possible
# once the Methionine/TSS is located downstream of the
# OMSR max site
if right_window_size <= 0:
right_window_size = stg._TCODE_3P_WINDOWSIZE
# calculate the average TCODE scores for the windows
( tcode5p,tcode3p ) = theorf.tcode_entropy_of_pos(
aaPos,
window_left=stg._TCODE_5P_WINDOWSIZE,
window_right=right_window_size,
)
stg._tcode5pscore[(org,orfid,aaPos,dnaPos)] = tcode5p
stg._tcode3pscore[(org,orfid,aaPos,dnaPos)] = tcode3p
# return filled tsscollection graph
return stg
def ExonCollectionGraph2TranslationalStartSiteCollectionGraph(gra):
"""
Convert ECG -> TranslationalStartSiteCollectionGraph
@attention: only in use when ECG is a FirstExon ECG
@rtype: TranslationalStartSiteCollectionGraph
@return: TranslationalStartSiteCollectionGraph instance to be placed in the CBG
"""
newgra = TranslationalStartSiteCollectionGraph()
newgra.ALIGNED_SITE_AA_OFFSET = 10
newgra.MIN_PSSM_SCORE = -0.0
for node in gra.get_nodes():
# exon node is ( org, orf, ntpos), TSS node ( org, orf, aapos, ntpos )
newnode = ( node[0], node[1], node[2]/3, node[2] )
tss = gra._node_object[node].acceptor
newgra.add_node_and_object(newnode,tss)
newgra._node_pssm[newnode] = tss.pssm_score
for nodeA,nodeB in newgra.pairwisecrosscombinations_node():
entropyQ = 1.0
entropyS = 1.0
newgra.add_edge(nodeA,nodeB,wt=1.0)
newgra._edge_binary_entropies[(nodeA,nodeB)] = (entropyQ,entropyS)
newgra._edge_binary_entropies[(nodeB,nodeA)] = (entropyS,entropyQ)
# Get tcode data for these start codon nodes
# Assuming that this is indeed the start-codon,
for (org,orfid,aaPos,dnaPos) in newgra.get_nodes():
theorf = gra.get_orfs_of_graph(organism=org)[0]
# calculate the average TCODE scores for the windows
( tcode5p,tcode3p ) = theorf.tcode_entropy_of_pos(
aaPos,
window_left=newgra._TCODE_5P_WINDOWSIZE,
window_right=newgra._TCODE_3P_WINDOWSIZE,
)
newgra._tcode5pscore[(org,orfid,aaPos,dnaPos)] = tcode5p
newgra._tcode3pscore[(org,orfid,aaPos,dnaPos)] = tcode3p
# return the TranslationalStartSiteCollectionGraph
return newgra
def harvest_elegiable_tss_sites(self,max_aa_distance=ALIGNED_TSS_MAX_AA_DISTANCE,
tss_min_pssm_score=TSS_MIN_PSSM_SCORE,
skip_nonelegiable_sites=True):
"""
"""
# update minimal pssm score to stg collection object
stg = TranslationalStartSiteCollectionGraph()
stg.MIN_PSSM_SCORE = tss_min_pssm_score
stg.ALIGNED_SITE_AA_OFFSET = max_aa_distance
# First, proces each individual organism.
for org in self.organism_set():
# take the first (and only) orf of this organism
theorf = self.get_orfs_of_graph(organism=org)[0]
# ready if there are no potential tss loci (no ATG sequence)
if not theorf.has_start(): continue
# scan for tss loci
theorf.scan_orf_for_pssm_tss(min_pssm_score=tss_min_pssm_score)
if skip_nonelegiable_sites:
# get the considered TSS range
(min_aa_pos, min_nt_pos) = self.minimal_eligable_tss_position(org)
(max_aa_pos, max_nt_pos) = self.maximal_eligable_tss_position(org)
else:
(min_aa_pos, min_nt_pos) = None, None
(max_aa_pos, max_nt_pos) = None, None
for tss in theorf._tss_sites:
# check if we can ignore this site
if min_nt_pos and tss.pos < min_nt_pos: continue
if max_nt_pos and tss.pos > max_nt_pos: continue
# an accepted site; add to TSS Collection Graph
startpos = tss.pos / 3
tssNode = ( org, theorf.id, startpos, tss.pos )
stg.add_node_and_object(tssNode,tss)
# Second, evaluate all cross combinations
for ( (a,b,c,d),(g1,o1),(g2,o2) ), pacbporf in self.pacbps.iteritems():
# only proces this combination if both organisms have splice sites!
if g1 not in stg.organism_set(): continue
if g2 not in stg.organism_set(): continue
# now loop over all TSS in Query and Sbjct
# and align them in a graph; an edge is added if 2 sites
# are less then ``max_aa_distance`` apart from each other
for tssQ in stg.get_organism_objects(g1):
# the node that represents this site
startQpos = tssQ.pos / 3
startQnode = ( g1, o1, startQpos, tssQ.pos )
for tssS in stg.get_organism_objects(g2):
# the node that represents this site
startSpos = tssS.pos / 3
startSnode = ( g2, o2, startSpos, tssS.pos )
# get distance between (aligned) start-codons
dist = pacbporf.get_distance_aligned_protein_positions(
query=startQpos,sbjct=startSpos)
# continue if distance between start sites is to big
if dist > max_aa_distance: continue
# calculate binary entropies from both positions
startQpositionPos,phaseQ = pacbporf.dnaposition_query(tssQ.pos,forced_return=True)
startSpositionPos,phaseS = pacbporf.dnaposition_sbjct(tssS.pos,forced_return=True)
entropyQ = pacbporf.alignment_entropy(startQpositionPos,method='left')
entropyS = pacbporf.alignment_entropy(startSpositionPos,method='left')
# calculate a weight from distance between startQpos and startSpos
wt = 1.0 / ( 1.0 + float(dist) )
# check if edge already in graph
if stg.has_edge( startQnode, startSnode ):
_wt = stg.weights[( startQnode, startSnode )]
if wt > _wt:
stg.set_edge_weight( startQnode, startSnode, wt=wt )
# and add binary entropy values
stg._edge_binary_entropies[(startQnode, startSnode)] = (entropyQ,entropyS)
stg._edge_binary_entropies[(startSnode, startQnode)] = (entropyS,entropyQ)
else:
stg.add_edge( startQnode, startSnode, wt=wt )
# and add binary entropy values
stg._edge_binary_entropies[(startQnode, startSnode)] = (entropyQ,entropyS)
stg._edge_binary_entropies[(startSnode, startQnode)] = (entropyS,entropyQ)
# Get tcode data for these start codon nodes
# Assuming that this is indeed the start-codon,
# the stretch of ATG untill max(OMSR) will be coding.
# Take the length of this stretch (in nt) as right/3p/upstream window size
omsr = self.overall_minimal_spanning_range()
for (org,orfid,aaPos,dnaPos) in stg.get_nodes():
theorf = self.get_orfs_of_graph(organism=org)[0]
right_window_size = ( max(omsr[(org,orfid)])+1 - aaPos )*3
# confirm that window size is not < 0; this is possible
# once the Methionine/TSS is located downstream of the
# OMSR max site
if right_window_size <= 0:
right_window_size = stg._TCODE_3P_WINDOWSIZE
# calculate the average TCODE scores for the windows
( tcode5p,tcode3p ) = theorf.tcode_entropy_of_pos(
aaPos,
window_left=stg._TCODE_5P_WINDOWSIZE,
window_right=right_window_size,
)
stg._tcode5pscore[(org,orfid,aaPos,dnaPos)] = tcode5p
stg._tcode3pscore[(org,orfid,aaPos,dnaPos)] = tcode3p
# return filled tsscollection graph
return stg
