def parseALERecFile(nfrec, reftreelen=None, restrictclade=None, skipEventFreq=False, skipLines=False, nsample=[], returnDict=False):
line = ''
lrecgt = []
restrictlabs = []
frec = open(nfrec, 'r')
while not line.startswith('S:\t'):
line = frec.readline()
# extract node labels from reconciled species tree
spetree = tree2.AnnotatedNode(nwk=line.strip('\n').split('\t')[1], namesAsNum=True)
spetree.complete_node_ids()
if reftreelen:
if not spetree.hasSameTopology(reftreelen): raise IndexError, "reference tree from $2 has not the same topology as that extracted from reconciliation output file $1"
for node in spetree:
# extract branch length from topologically identical tree from $2
matchclade = reftreelen.map_to_node(node.get_leaf_labels())
node.set_lg(matchclade.lg())
if restrictclade:
for restrictnodelab in restrictclade.split(','):
restrictlabs += spetree[restrictnodelab].get_children_labels()
subspetree = spetree.restrictToLeaves(restrictlabs, force=True)
else:
subspetree = spetree
while not line.endswith('reconciled G-s:\n'):
line = frec.readline()
for i in range(2): line = frec.readline() # skips 2 lines
# extract reconciled gene tree(s)
recgtlines = []
k = 0
while not line.startswith('#'):
if (not nsample) or (k in nsample):
if not skipLines: recgtlines.append(line)
rectree = tree2.AnnotatedNode(nwk=line.strip('\n'), namesAsNum=True)
rectree.complete_node_ids()
lrecgt.append(rectree)
line = frec.readline()
k += 1
dnodeevt = {}
if not skipEventFreq:
for i in range(3): line = frec.readline() # skips 3 lines
# extract node-wise event frequency / copy number info
for line in frec:
if line=='\n': continue
lsp = line.strip('\n').split('\t')
dnodeevt[lsp[1]] = [float(s) for s in lsp[2:]]
frec.close()
if returnDict:
return {'spetree':spetree, 'subspetree':subspetree, 'lrecgt':lrecgt, 'recgtlines':recgtlines, 'restrictlabs':restrictlabs, 'dnodeevt':dnodeevt}
else:
return [spetree, subspetree, lrecgt, recgtlines, restrictlabs, dnodeevt]
def loadRefPopTree(nfrefspetree, nfpop):
# annotate reference species tree with ancestral population node labels
lnamepops = []
with open(nfpop, 'r') as fpop:
for line in fpop:
if line.startswith('#'): continue
lsp = line.rstrip('\n').split('\t')
lnamepops.append((lsp[0], tuple(lsp[1].split())))
refspetree = tree2.AnnotatedNode(file=nfrefspetree)
refspetree.complete_internal_labels(order=0, ffel=True)
refspetree.complete_node_ids(force=True)
annotatePopulationInSpeciesTree(refspetree, lnamepops, returnCopy=False, returnAncNodes=False)
dspe2pop = getdspe2pop(lnamepops)
nfrefspetreeout = nfrefspetree.rsplit('.', 1)[0]+'_internalPopulations.nwk'
refspetree.write_newick(nfrefspetreeout, ignoreBS=True)
return (refspetree, dspe2pop)
return [field.strip('" ') for field in line.rstrip('\n').split('\t')]
if len(sys.argv)<3:
print "Usage: %s /path/to/lineage_module_event_table /path/to/reference_tree /path/to/output_folder [/path/to/reconciliation_folder]"
sys.exit(2)
nflnflineagecommevents = sys.argv[1]
nfrefspetree = sys.argv[2]
dirout = sys.argv[3]
# optional argument to give context of gene family occurrence to gene lineage
if len(sys.argv)>4:
dirrec = sys.argv[4]
else:
dirrec = None
refspetree = tree2.AnnotatedNode(file=nfrefspetree)
with open(nflnflineagecommevents, 'r') as flnflineagecommevents:
lnflineagecommevents = [line.rstrip('\n') for line in flnflineagecommevents]
dfamspetree = {}
for nflineagecommevents in lnflineagecommevents:
flineagecommevents = open(nflineagecommevents, 'r')
lineagecomm = os.path.basename(nflineagecommevents).rsplit('.', 1)[0]
dirlineageout = os.path.join(dirout, lineagecomm)
if not os.path.isdir(dirlineageout): os.mkdir(dirlineageout)
curfamily = None
curlineage = None
curspetree = None
dnodefreq = {}
ltrans = []
def main(nfgenetree, diraln, dirout, outtag, mkdircons=True, **kw):
aliformatin = kw.get('aliformatin')
diridentseq = kw.get('diridentseq')
isparallel = kw.get('isparallel')
print nfgenetree
bnspl = os.path.basename(nfgenetree).split('.')
if bnspl[0].startswith('RAxML_'): bngt = bnspl[1]
else: bngt = bnspl[0]
globaln = "%s/%s*%s*" % (diraln, bngt, aliformatin[:3])
try:
nfaln = glob.glob(globaln)[0]
except IndexError:
globaln = "%s/%s*aln*" % (diraln, bngt)
nfaln = glob.glob(globaln)[0]
print nfaln
bnfaln = os.path.basename(nfaln).split('.')[0]
if bnspl[0].startswith('RAxML_rootedTree'):
# tree is already rooted, but the branch supports are storred in the comments
genetree = tree2.AnnotatedNode(file=nfgenetree, keep_comments=True)
for n in genetree:
if str(n.comment()).isdigit():
n.set_bs(float(n.comment()))
else:
# trees is unrooted
genetree = tree2.read_check_newick(nfgenetree,
treeclass='AnnotatedNode')
# tree is interpreted here as trifurcated at the root ; root it.
genetree.resolveNode(outgroups='subroot')
# there will be deepcopy operation on the tree, either to save its state before pruning (pop) below or in select_clades.
# a deepcopy operation on a recursive tree2.Node object induces a cycle of ~7 function calls per nested node
# knowing that there are (2*n)-1 nodes in a tree (n being the nuber of tree leaves),
# one should set the recursion limit >> 7*2*n ; on sequential calls, set it to 10*2*n to be on the safe side with overheads of higher level function calls.
if not isparallel:
adddepth(currentmaxreccursdepths, bngt, 10 * 2 * genetree.nb_leaves())
# deal with potential information on sets of identical sequences
didseq = {}
if diridentseq:
# parse pairs of (reference, redundant) sequences that are identical
globidseq = "%s/*%s*" % (diridentseq, bngt)
gnfidseq = glob.glob(globidseq)
if not gnfidseq:
raise OSError, "cannot find file matching pattern: '%s'" % globidseq
nfidseq = gnfidseq[0]
with open(nfidseq, 'r') as fidseq:
for line in fidseq:
refidseq, redidseq = line.rstrip('\n').split('\t')
didseq.setdefault(refidseq, []).append(redidseq)
if didseq:
# remove any redundant sequence from the gene tree before processing
cleangenetree = copy.deepcopy(genetree)
gtleaves = set(cleangenetree.get_leaf_labels())
for refidseq, redidseqs in didseq.iteritems():
for redidseq in redidseqs:
if redidseq in gtleaves:
cleangenetree.pop(redidseq)
gtleaves.remove(redidseq)
else:
cleangenetree = genetree
else:
cleangenetree = genetree
# detect unresolved clades
constraintswithsingles = mark_unresolved_clades(
cleangenetree, **kw) #, pruneSelected=True, inclusive=True
if verbose: print 'constraintswithsingles =', constraintswithsingles
# add to identical sequence map to the constrained clades definitions
newconstraintsfromidseqs = []
for refidseq, redidseqs in didseq.iteritems():
# scan for existing clade that would contain the reference
for c in constraintswithsingles:
if refidseq in c:
c += redidseqs
break
else:
newconstraintsfromidseqs.append([refidseq] + redidseqs)
if verbose: print 'newconstraintsfromidseqs =', newconstraintsfromidseqs
# for reporting, filter out contraint clades that are just made of one leaf (NB: these are useful for proper definitition of other constraint clades, when nested, non-inclusive clades are allowed)
constraints = [
c for c in constraintswithsingles + newconstraintsfromidseqs
if len(c) > 1
]
if verbose: print 'constraints =', constraints
# write out subalignments and the main alignment with collapsed clades
loutgroups = restrict_alignment_representative_leaves(constraints,
genetree,
nfaln,
dirout,
radout=bnfaln,
selectRepr=0,
didseq=didseq,
**kw)
if not 'mbc' in supressout:
mbcoutd, mbcext = doutdext['mbc']
if mkdircons is True:
mbcoutd = os.path.join(mbcoutd, bnfaln)
dout = os.path.join(dirout, mbcoutd)
if not os.path.isdir(dout):
os.mkdir(dout)
for i, constraint in enumerate(constraints):
cladename = "clade%d" % i
# write out MrBayes clade constraint for the sub-alignment, in order to compute subalignment samples and/or ancestral sequence
write_out_MrBayes_clade_constraints(
[constraint],
loutgroups[i],
os.path.join(dirout, mbcoutd,
bnfaln + '-' + cladename + '.' + mbcext),
ilist=[i],
verbose=verbose)
if not 'cgt' in supressout:
fmtcoltree = kw.get('format_color_tree')
cgtoutd, cgtext = doutdext['cgt']
colour_tree_with_constrained_clades(genetree, constraints, force=True)
genetree.complete_internal_labels(order=0, ffel=True)
if fmtcoltree.lower() in ['xml', 'phyloxml']:
genetree.write_phyloXML(os.path.join(dirout, cgtoutd,
bnfaln + '-%s.xml' % cgtext),
ignoreBS=True)
elif fmtcoltree.lower() in ['nex', 'nexus']:
genetree.write_nexus(os.path.join(dirout, cgtoutd,
bnfaln + '-%s.nex' % cgtext),
ignoreBS=True)
else:
raise ValueError, "specified format '%s' for output coloured-branch tree is not valid; please select among '[phylo]xml' or 'nex[us]'" % fmtcoltree
# done risking going over reccursion limt
if not isparallel: rmdepth(currentmaxreccursdepths, bngt)
def main(nfrec,
nfreftree,
nfgenetree,
maxrecgt=1,
recformat='tera',
sgsep='_',
phylofact=1000.0,
restrictclade=None,
verbose=False,
**kw):
try:
genetree = tree2.Node(file=nfgenetree, namesAsNum=True)
except ValueError:
genetree = tree2.Node(file=nfgenetree,
namesAsNum=True,
branch_lengths=False)
reftree = tree2.AnnotatedNode(file=nfreftree, namesAsNum=True)
if restrictclade: st = reftree.restrictToLeaves(restrictclade)
else: st = reftree
# check presence of outgroup/dead lineage branch if necessary
if recformat == 'tera':
if not (kw.get('noDeadStories') or
(deadlabnum in st.get_leaf_labels())):
if (outtaxlab in st.get_leaf_labels()):
# must adapt mowgli-compliant species tree
st[outtaxlab].edit_label(deadlabnum)
else:
maxd = reftree.max_leaf_distance()
outgroup = tree2.AnnotatedNode(lleaves=[deadlabnum])
outgroup.get_children()[0].set_lg(maxd * 3)
outgroup.link_child(reftree, newlen=maxd * 2)
reftree = outgroup
reftree.complete_internal_labels(prefix='')
# else:
# raise ValueError, "the provided species tree should feature a branch labaelled 'OUTGROUP' or '-1' to represent the dead/unsampled lineages"
elif recformat == 'mowgli':
if not (outtaxlab in st.get_leaf_labels()):
if (deadlabnum in st.get_leaf_labels()):
# must adapt mowgli-compliant species tree
st[deadlabnum].edit_label(outtaxlab)
else:
outgroup = tree2.AnnotatedNode(lleaves=[outtaxlab])
outgroup.get_children()[0].set_lg(maxd * 3)
outgroup.link_child(reftree, newlen=maxd * 2)
reftree = outgroup
reftree.complete_internal_labels(prefix='')
# else:
# raise ValueError, "the provided species tree should feature a branch labaelled 'OUTGROUP' or '-1' to represent the dead/unsampled lineages"
for i, rec in enumerate(
parseTERARecFile(nfrec,
genetree=genetree,
recformat=recformat,
sgsep=sgsep,
verbose=verbose,
**kw)):
dnodefreq, dlevt = rec
# write SVG species tree
tag = '_no_dead' if kw.get('noDeadStories') else ''
nfoutspe = '%s_%d_maprec2spetree%s.svg' % (nfrec, i, tag)
lleaffreq = [(lab, f) for lab, f in dnodefreq.items()
if st[lab].is_leaf()]
st.writeSvgTree(nfoutspe, padleaves=True, supports=False, phylofact=phylofact, branchwidths=dnodefreq, textorbit=5, \
treetype='species', transfers=dlevt['T'], duplications=dlevt['D'], losses=dlevt['L'], counts=lleaffreq, \
transferwidth='freq', modstyle="stroke-width:1; ", padstyle="stroke:red; stroke-width:0.5; stroke-dasharray:1,1; ")
# transfercolor='green',
print os.path.basename(nfoutspe)