def remove_kink(node, curroot):
"""
smooth the kink created by prunning
to prevent creating orphaned tips
after prunning twice at the same node
"""
if node == curroot:
print "fix bifurcating root by moving the root away to an adjacent none-tip"
if curroot.nchildren == 1:
curroot = curroot.children[0]
curroot.parent = None
assert curroot.nchildren == 2, \
"check tree root format: "+newick3.tosting(curroot)+";"
if curroot.children[0].istip: #the other child is not tip
curroot = phylo3.reroot(curroot, curroot.children[1])
else:
curroot = phylo3.reroot(curroot, curroot.children[0])
else:
#---node---< all nodes should have one child only now
length = node.length + (node.children[0]).length
par = node.parent
kink = node
node = node.children[0]
#parent--kink---node<
par.remove_child(kink)
par.add_child(node)
node.length = length
return node, curroot
def remove_kink(node,curroot): """ smooth the kink created by prunning to prevent creating orphaned tips after prunning twice at the same node """ if node == curroot: print "fix bifurcating root by moving the root away to an adjacent none-tip" if curroot.nchildren == 1: curroot = curroot.children[0] curroot.parent = None assert curroot.nchildren == 2, \ "check tree root format: "+newick3.tosting(curroot)+";" if curroot.children[0].istip: #the other child is not tip curroot = phylo3.reroot(curroot,curroot.children[1]) else: curroot = phylo3.reroot(curroot,curroot.children[0]) else: #---node---< all nodes should have one child only now length = node.length + (node.children[0]).length par = node.parent kink = node node = node.children[0] #parent--kink---node< par.remove_child(kink) par.add_child(node) node.length = length return node,curroot
def remove_kink(node,curroot): if node == curroot and curroot.nchildren == 2: #move the root away to an adjacent none-tip if curroot.children[0].istip: #the other child is not tip curroot = phylo3.reroot(curroot,curroot.children[1]) else: curroot = phylo3.reroot(curroot,curroot.children[0]) #---node---< all nodes should have one child only now length = node.length + (node.children[0]).length par = node.parent kink = node node = node.children[0] #parent--kink---node< par.remove_child(kink) par.add_child(node) node.length = length return node,curroot
def remove_kink(node,curroot): if node == curroot and curroot.nchildren == 2: #move the root away to an adjacent none-tip if curroot.children[0].istip: #the other child is not tip curroot = phylo3.reroot(curroot,curroot.children[1]) else: curroot = phylo3.reroot(curroot,curroot.children[0]) #---node---< all nodes should have one child only now length = node.length + (node.children[0]).length par = node.parent kink = node node = node.children[0] #parent--kink---node< par.remove_child(kink) par.add_child(node) node.length = length return node,curroot
def extract_rooted_ingroup_clades(root, ingroups, outgroups, min_ingroup_taxa):
"""
input a tree with ingroups and at least 1 outgroups
output a list of rooted ingroup clades
"""
inclades = []
while True:
max_score, direction, max_node = 0, "", None
for node in root.iternodes():
front, back = 0, 0
front_names_set = set(get_front_names(node))
for name in front_names_set:
if name in outgroups:
front = -1
break
elif name in ingroups:
front += 1
else:
sys.exit("Check taxonID " + name)
back_names_set = set(get_back_names(node, root))
for name in back_names_set:
if name in outgroups:
back = -1
break
elif name in ingroups:
back += 1
else:
sys.exit("Check taxonID " + name)
if front > max_score:
max_score, direction, max_node = front, "front", node
if back > max_score:
max_score, direction, max_node = back, "back", node
#print max_score,direction
if max_score >= min_ingroup_taxa:
if direction == "front":
inclades.append(max_node)
kink = max_node.prune()
if len(root.leaves()) > 3:
newnode, root = remove_kink(kink, root)
else:
break
elif direction == "back":
par = max_node.parent
par.remove_child(max_node)
max_node.prune()
inclades.append(phylo3.reroot(root, par)) #flip dirction
if len(max_node.leaves()) > 3:
max_node, root = remove_kink(max_node, max_node)
else:
break
else:
break
return inclades
def remove_kink(node,curroot): """ smooth the kink created by prunning to prevent creating orphaned tips after prunning twice at the same node """ if node == curroot and curroot.nchildren == 2: #move the root away to an adjacent none-tip if curroot.children[0].istip: #the other child is not tip curroot = phylo3.reroot(curroot,curroot.children[1]) else: curroot = phylo3.reroot(curroot,curroot.children[0]) #---node---< all nodes should have one child only now length = node.length + (node.children[0]).length par = node.parent kink = node node = node.children[0] #parent--kink---node< par.remove_child(kink) par.add_child(node) node.length = length return node,curroot
def cut_long_branches(curroot, cutoff):
going = True
subtrees = [] #store all subtrees after cutting
if curroot.nchildren == 2: #fix the root
#move the root away to an adjacent none-tip internal node
if curroot.children[0].istip: #the other child is not tip
curroot = phylo3.reroot(curroot, curroot.children[1])
else: #tree has >=4 leaves so the other node cannot be tip
curroot = phylo3.reroot(curroot, curroot.children[0])
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 != curroot and node.length > cutoff:
subtrees.append(node)
node = node.prune()
if len(curroot.leaves()) >= 4:
node, curroot = remove_kink(node, curroot)
going = True
break
subtrees.append(curroot) #write out the residue after cutting
return subtrees
def cut_long_branches(curroot,cutoff): going = True subtrees = [] #store all subtrees after cutting if curroot.nchildren == 2: #fix the root #move the root away to an adjacent none-tip internal node if curroot.children[0].istip: #the other child is not tip curroot = phylo3.reroot(curroot,curroot.children[1]) else: #tree has >=4 leaves so the other node cannot be tip curroot = phylo3.reroot(curroot,curroot.children[0]) 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 != curroot and node.length > cutoff: subtrees.append(node) node = node.prune() if len(curroot.leaves()) >= 4: node,curroot = remove_kink(node,curroot) going = True break subtrees.append(curroot) #write out the residue after cutting return subtrees
def reroot_with_monophyletic_outgroups(root):
lvs = root.leaves()
outgroup_matches = {} #key is label, value is the tip node object
#Since no taxon repeat in outgroups name and leaf is one-to-one
outgroup_labels = []
for leaf in lvs:
label = leaf.label
name = get_name(label)
if name in OUTGROUPS:
outgroup_matches[label] = leaf
outgroup_labels.append(leaf.label)
if len(outgroup_labels) == 1: #one single outgroup
#cannot reroot on a tip so have to go one more node into the ingroup
new_root = outgroup_matches[outgroup_labels[0]].parent
return phylo3.reroot(root,new_root)
else: #has multiple outgroups. Check monophyly and reroot
newroot = None
for node in root.iternodes():
if node == root: continue #skip the root
front_names = get_front_names(node)
back_names = get_back_names(node,root)
front_in_names,front_out_names,back_in_names,back_out_names = 0,0,0,0
for i in front_names:
if i in OUTGROUPS: front_out_names += 1
else: front_in_names += 1
for j in back_names:
if j in OUTGROUPS: back_out_names += 1
else: back_in_names += 1
if front_in_names==0 and front_out_names>0 and back_in_names>0 and back_out_names==0:
newroot = node #ingroup at back, outgroup in front
break
if front_in_names>0 and front_out_names==0 and back_in_names==0 and back_out_names>0:
newroot = node.parent #ingroup in front, outgroup at back
break
if newroot != None:
return phylo3.reroot(root,newroot)
else: return None
def extract_rooted_ingroup_clades(root,ingroups,outgroups,min_ingroup_taxa):
"""
input a tree with ingroups and at least 1 outgroups
output a list of rooted ingroup clades
"""
inclades = []
while True:
max_score,direction,max_node = 0,"",None
for node in root.iternodes():
front,back = 0,0
front_names_set = set(get_front_names(node))
for name in front_names_set:
if name in outgroups:
front = -1
break
elif name in ingroups: front += 1
else: sys.exit("Check taxonID "+name)
back_names_set = set(get_back_names(node,root))
for name in back_names_set:
if name in outgroups:
back = -1
break
elif name in ingroups: back += 1
else: sys.exit("Check taxonID "+name)
if front > max_score:
max_score,direction,max_node = front,"front",node
if back > max_score:
max_score,direction,max_node = back,"back",node
#print max_score,direction
if max_score >= min_ingroup_taxa:
if direction == "front":
inclades.append(max_node)
kink = max_node.prune()
if len(root.leaves()) > 3:
newnode,root = remove_kink(kink,root)
else: break
elif direction == "back":
par = max_node.parent
par.remove_child(max_node)
max_node.prune()
inclades.append(phylo3.reroot(root,par))#flip dirction
if len(max_node.leaves()) > 3:
max_node,root = remove_kink(max_node,max_node)
else: break
else: break
return inclades
import phylo3, newick3
import sys
if len(sys.argv) < 3:
print __doc__
print "usage: roottrees <treesfile> <outgroupsfile>"
sys.exit(0)
treesfname = sys.argv[1]
outgroupsfname = sys.argv[2]
treesfile = open(treesfname,"r")
outgroupsfile = open(outgroupsfname,"r")
outgroup_names = [line.strip() for line in outgroupsfile.readlines()]
rooted_trees = []
for line in treesfile:
tree = newick3.parse(line)
outgroup = phylo3.getMRCA(tree, outgroup_names)
rooted_tree = phylo3.reroot(tree, outgroup)
rooted_trees.append(rooted_tree)
outfile = open(treesfname.rsplit(".tre",1)[0]+".rooted.tre","w")
for tree in rooted_trees:
outfile.write(newick3.to_string(tree)+";\n")