sumlabels? about dendropy

[jeetsukumaran]

Unfortunately, the script has not yet been updated for DendroPy 4. It *may* work with a little modification directly, or it may not. And yes, your suggestion for a null value makes perfect sense. I actually responded to someone else with a script that definitely works for DendroPy 4 to do what you want (albeit, without proper handling for the N/A split support). If you are approaching this with some Python/scripting background, it should be easy enough to tweak to accommodate your situation. Otherwise, let me know and I will see if I can provide a new version of sumlabels ... though I cannot promise you a timeline for this! ~~~ #! /usr/bin/env python import dendropy # our common namespace tns = dendropy.TaxonNamespace() # load the trees t1 = dendropy.Tree.get( path="maxpars.tre", schema="nexus", taxon_namespace=tns, rooting="force-rooted", # or "force-unrooted" if trees unrooted ) t2 = dendropy.Tree.get( path="raxml.tre", schema="newick", taxon_namespace=tns, rooting="force-rooted", # or "force-unrooted" if trees unrooted ) t3 = dendropy.Tree.get( path="mrbayes.tre", schema="nexus", taxon_namespace=tns, rooting="force-rooted", # or "force-unrooted" if trees unrooted ) # calculate the split hashes so that we can correlate splits # across trees t1.encode_bipartitions() t2.encode_bipartitions() t3.encode_bipartitions() # for each internal node in tree 3 for nd in t3.preorder_internal_node_iter(): # check/skip root node if nd is t3.seed_node: continue # set up label parts label_parts = [] # look up edge corresponding to this node # on tree1 edge = t1.bipartition_edge_map[nd.edge.bipartition] # (if support on tree1 is represented by branch length) # label_parts.append(str(edge.length)) label_parts.append(edge.head_node.label) # look up edge corresponding to this node # on tree2 edge = t2.bipartition_edge_map[nd.edge.bipartition] # support on tree2 is represented by internal node label label_parts.append(edge.head_node.label) # if support on tree3 is given by metadata # label_parts.append(nd.annotations["prob"].value) label_parts.append(nd.label) # print(label_parts) # assign label nd.label = "/".join(label_parts) # output the tree print(t3.as_string("newick")) ~~~

…

On 11/24/16 4:08 AM, courtneystairs wrote: Hi there! A couple years ago I was using sumlabels.py to map support values of multiple different onto one tree. I see that the script has not been updated in some time and I cannot seem to find it in the newest repository. Can you tell me if it has been incorporated into another package or with another name? I also have a specific question about the script. I noticed that if it does not find the same support value for a given bipartition, it will not map a value to the output tree. This makes sense since that bipartition was not found! However, when mapping the values, its hard to determine which value was added from which tree. Is there a way to get the script to write a '-' or 'nf.' in place of nothing? Thank you! Courtney — You are receiving this because you are subscribed to this thread. Reply to this email directly, view it on GitHub <#71>, or mute the thread <https://github.com/notifications/unsubscribe-auth/AABmR6tmKifUFmuukJMxG4GsnRqh7s7Tks5rBVQogaJpZM4K7a6l>.

--

-------------------------------------- Jeet Sukumaran -------------------------------------- [email protected] -------------------------------------- Blog/Personal Pages: http://jeetworks.org/ GitHub Repositories: http://github.com/jeetsukumaran Photographs (as stream): http://www.flickr.com/photos/jeetsukumaran/ Photographs (by galleries): http://www.flickr.com/photos/jeetsukumaran/sets/ --------------------------------------

[courtneystairs]

Great!! Thank you. I'll /try/ and tweak this script this morning and post what I get working. Thank you!!

[courtneystairs]

Thanks again for the script. It differs a little from sumlabels in that it threw an error if it didn't find the bipartition (whereas sumlabels would just not write anything, not even a delimiter). To combat that I have added 2 try/except clauses and command line argument functionality with arguments. One day I'll make it prettier, but for now this works:

python sumlabel_DP4.py [tree to map to] [tree] [tree]

it will write 'nf' on the branch if it didn't find the bipartition.

#! /usr/bin/env python
import sys
import dendropy


# usage: python sumlabel_DP4.py [tree to map to] [tree] [tree] 
# our common namespace
tns = dendropy.TaxonNamespace()


# load the trees
t1 = dendropy.Tree.get(
     path=sys.argv[3],
     schema="newick",
     taxon_namespace=tns,
     rooting="force-unrooted")
t2 = dendropy.Tree.get(
     path=sys.argv[2],
     schema="newick",
     taxon_namespace=tns,
     rooting="force-unrooted")
t3 = dendropy.Tree.get(
     path=sys.argv[1],
     schema="newick",
     taxon_namespace=tns,rooting="force-unrooted")

# calculate the split hashes so that we can correlate splits
# across trees
t1.encode_bipartitions()
t2.encode_bipartitions()
t3.encode_bipartitions()

# for each internal node in tree 3
for nd in t3.preorder_internal_node_iter():

     # check/skip root node
     if nd is t3.seed_node:
         continue

     # set up label parts
     label_parts = []

     # look up edge corresponding to this node
     # on tree1 if not found, write 'nf'
     try:
     	edge = t1.bipartition_edge_map[nd.edge.bipartition]
	label_parts.append(edge.head_node.label)

     except KeyError:
        label_parts.append('nf')
     # (if support on tree1 is represented by branch length)
     # label_parts.append(str(edge.length))
     # look up edge corresponding to this node
     # on tree2 if  not found, write 'nf'
     try:
	edge = t2.bipartition_edge_map[nd.edge.bipartition]
	label_parts.append(edge.head_node.label)
     # support on tree2 is represented by internal node label
     except KeyError:
	label_parts.append('nf')
     # if support on tree3 is given by metadata
     # label_parts.append(nd.annotations["prob"].value)
     label_parts.append(nd.label)
     # print(label_parts)

     # assign label
     nd.label = "/".join(label_parts)
# output the tree
print(t3.as_string("newick"))

[jeetsukumaran]

The original application has been ported in DendroPy 4 here: 86e77ec .

May tweak the UI before formal release, but it is workable as it stands now.