def cut_long_internal_branches(curroot, cutoff):
"""cut long branches and output all subtrees with at least 4 tips"""
going = True
subtrees = [] #store all subtrees after cutting
while going:
going = False #only keep going if long branches were found during last round
for node in curroot.iternodes(): #Walk through nodes
if node.istip or node == curroot: continue
if node.nchildren == 1:
node, curroot = remove_kink(node, curroot)
going = True
break
child0, child1 = node.children[0], node.children[1]
if node.length > cutoff:
print node.length
if not child0.istip and not child1.istip and child0.length + child1.length > cutoff:
print child0.length + child1.length
if count_taxa(child0) >= 4:
subtrees.append(child0)
if count_taxa(child1) >= 4:
subtrees.append(child1)
else:
subtrees.append(node)
node = node.prune()
if len(curroot.leaves()) > 2: #no kink if only two left
node, curroot = remove_kink(node, curroot)
going = True
break
if count_taxa(curroot) >= 4:
subtrees.append(curroot) #write out the residue after cutting
return subtrees
def prune(score_tuple,node,root,pp_trees): if score_tuple[0] > score_tuple[1]: #prune front print "prune front" pp_trees.append(node) par = node.prune() if par != None and len(root.leaves()) >= 3: par,root = tree_utils.remove_kink(par,root) return root,node == root else: if node != root: #prune back par = node.parent #par--node< par.remove_child(node) if par.parent != None: par,root = tree_utils.remove_kink(par,root) node.prune() print "prune back" pp_trees.append(root) if len(node.leaves()) >= 3: node,newroot = tree_utils.remove_kink(node,node) else: newroot = node return newroot,False #original root was cutoff, not done yet
def main(inDIR, file_ending, branch_len_cutoff, min_taxa, outDIR):
"""cut long branches and output subtrees as .subtre files
if uncut and nothing changed betwee .tre and .subtree
copy the original .tre file to the outdir"""
if inDIR[-1] != "/": inDIR += "/"
if outDIR[-1] != "/": outDIR += "/"
min_taxa = int(min_taxa)
filecount = 0
cutoff = float(branch_len_cutoff)
print "cutting branches longer than", cutoff
for i in os.listdir(inDIR):
if not i.endswith(file_ending): continue
print i
filecount += 1
with open(inDIR + i, "r") as infile: #only 1 tree in each file
intree = newick3.parse(infile.readline())
try:
with open(inDIR + i[:i.find(".tre")] + ".tre",
"r") as infile: #the original .tre
raw_tree_size = len(
get_front_labels(newick3.parse(infile.readline())))
except: # did not refine this round. Use the .tre.tt.mm tree
raw_tree_size = len(get_front_labels(intree))
num_taxa = count_taxa(intree)
if num_taxa < min_taxa:
print "Tree has", num_taxa, "less than", min_taxa, "taxa"
else:
print ".tre:", raw_tree_size, "tips; " + file_ending + ": " + str(
len(get_front_labels(intree))) + " tips"
subtrees = cut_long_internal_branches(intree, cutoff)
if len(subtrees) == 0:
print "No tree with at least", min_taxa, "taxa"
#elif raw_tree_size == len(subtrees[0].leaves()):
#copy(inDIR+i,outDIR+i)
#print "written to out directory unchanged"
else:
count = 0
outsizes = ""
for subtree in subtrees:
if count_taxa(subtree) >= min_taxa:
if subtree.nchildren == 2: #fix bifurcating roots from cutting
temp, subtree = remove_kink(subtree, subtree)
count += 1
with open(
outDIR + i.split(".")[0] + "_" + str(count) +
".subtree", "w") as outfile:
outfile.write(newick3.tostring(subtree) + ";\n")
outsizes += str(len(subtree.leaves())) + ", "
print count, "tree(s) wirtten. Sizes:", outsizes
assert filecount > 0, "No file end with " + file_ending + " in " + inDIR
def mask_monophyletic_tips(curroot,ignore=[]):
going = True
while going and curroot != None and len(curroot.leaves()) >= 4:
going = False
for node in curroot.iternodes(): # walk through nodes
if not node.istip: continue # only look at tips
name = get_name(node.label).split("_")[1]
for sister in node.get_sisters():
if sister.istip and name==get_name(sister.label).split("_")[1]: # mask
node = sister.prune()
if len(curroot.leaves()) >= 4:
if (node==curroot and node.nchildren==2) or (node!=curroot and node.nchildren==1):
node,curroot = remove_kink(node,curroot)
going = True
break
return curroot
def mask_paraphyletic_tips(curroot,ignore=[]):
going = True
while going and curroot != None and len(curroot.leaves()) >= 4:
going = False
for node in curroot.iternodes(): #walk through nodes
if not node.istip: continue #only look at tips
name = get_name(node.label).split("_")[1]
parent = node.parent
if node == curroot or parent == curroot or parent == None:
continue #no paraphyletic tips for the root
for para in parent.get_sisters():
if para.istip and name==get_name(para.label).split("_")[1]: # mask
node = para.prune()
if len(curroot.leaves()) >= 4:
if (node==curroot and node.nchildren==2) or (node!=curroot and node.nchildren==1):
node,curroot = remove_kink(node,curroot)
going = True
break
return curroot
def main(inDIR,file_ending,branch_len_cutoff,min_taxa,outDIR,log):
"""cut long branches and output subtrees as .subtre files
if uncut and nothing changed betwee .tre and .subtree
copy the original .tre file to the outdir"""
if inDIR[-1] != "/": inDIR += "/"
min_taxa = int(min_taxa)
filecount = 0
cutoff = float(branch_len_cutoff)
print("cutting branches longer than",cutoff)
for i in os.listdir(inDIR):
if not i.endswith(file_ending): continue
#print i
filecount += 1
with open(inDIR+i,"r") as infile: #only 1 tree in each file
intree = tree_reader.read_tree_string(infile.readline())
try:
with open(inDIR+i[:i.find(".tre")]+".tre","r") as infile: #the original .tre
raw_tree_size = len(get_front_labels(tree_reader.read_tree_string(infile.readline())))
except: # did not refine this round. Use the .tre.tt.mm tree
raw_tree_size = len(get_front_labels(intree))
num_taxa = len(intree.leaves())
if num_taxa < min_taxa:
print("Tree has",num_taxa,"less than", min_taxa,"taxa")
else:
#print ".tre:",raw_tree_size,"tips; "+file_ending+": "+str(len(get_front_labels(intree)))+" tips"
subtrees = cut_long_internal_branches(intree,cutoff,min_taxa)
if len(subtrees) == 0:
print("No tree with at least", min_taxa, "taxa")
else:
count = 0
outsizes = ""
for subtree in subtrees:
if len(subtree.leaves()) >= min_taxa:
if len(subtree.children) == 2: #fix bifurcating roots from cutting
temp,subtree = remove_kink(subtree,subtree)
count += 1
outname = outDIR+"/"+i.split(".")[0]+"_"+str(count)+".subtree"
print(outname)
with open(outname,"w") as outfile:
outfile.write(subtree.get_newick_repr(True)+";\n")
outsizes += str(len(subtree.leaves()))+", "
print(count,"tree(s) written. Sizes:",outsizes)
def mask_monophyletic_tips(curroot, unamb_chrDICT):
going = True
while going and len(curroot.leaves()) >= 4:
going = False
for node in curroot.iternodes(): #walk through nodes
if not node.istip: continue #only look at tips
for sister in node.get_sisters():
if sister.istip and get_name(node.label) == get_name(
sister.label): #masking
#print node.label,unamb_chrDICT[node.label],sister.label,unamb_chrDICT[sister.label]
if unamb_chrDICT[node.label] > unamb_chrDICT[sister.label]:
node = sister.prune()
else:
node = node.prune()
if len(curroot.leaves()) >= 4:
if (node == curroot and node.nchildren == 2) or (
node != curroot and node.nchildren == 1):
node, curroot = remove_kink(node, curroot)
going = True
break
return curroot
def mask_paraphyletic_tips(curroot, unamb_chrDICT):
going = True
while going and len(curroot.leaves()) >= 4:
going = False
for node in curroot.iternodes(): #walk through nodes
if not node.istip: continue #only look at tips
parent = node.parent
if node == curroot or parent == curroot:
continue #no paraphyletic tips for the root
for para in parent.get_sisters():
if para.istip and get_name(node.label) == get_name(para.label):
if unamb_chrDICT[node.label] > unamb_chrDICT[para.label]:
node = para.prune()
else:
node = node.prune()
if len(curroot.leaves()) >= 4:
if (node == curroot and node.nchildren == 2) or (
node != curroot and node.nchildren == 1):
node, curroot = remove_kink(node, curroot)
going = True
break
return curroot