search

ansariazim / treeBreaker

Algorithm to divide a phylogenetic tree into segments based on phenotypes at the leaves of the tree
25 stars 3 forks source link

the layout of output file of treeBreaker #2

Closed lindechun closed 6 years ago

lindechun commented 6 years ago

i gave put the correct parameter “inputfile_tree inputfile_phenotype outputfile”，and No error was reported during operation, but the last line of output file of treeBreaker isn't contains the newick string of the tree

ansariazim commented 6 years ago

Hi there, Could you leave your sample tree file and the pheno file on here. I'll have a look to see what is wrong. Also did you test treeBreaker on the test data in the "testData" directory? Best. Azim

lindechun commented 6 years ago

Yeah, I have tested the test data in the 'testData' directory and the result is correct.

I upload my tree file and pheno file on here, and the command is that " treeBreaker 6829_ML.tree MDR 6829_treeBreaker.txt treeBreaker.R 6829_treeBreaker.txt 6829_treeBreaker"

Thank you for helping me check for mistakes.

treeBreaker_file.zip

ansariazim commented 6 years ago

Hi there, I ran treeBreaker on your files and it worked fine. Not sure what the problem is. Given that it works on the testdata, the most likely case is that your input tree is not formatted correctly. Maybe you could recompile treeBreaker. If you still have problems, I might be able to send you binaries compiled on osx.

Just for consistancy, some of your internal nodes have names (numeric names like 0.269) and some don't. It might make more sense to be consistant, ie either name all internal nodes or don't name any of them. The other thing is that numeric names like 0.269 might not be unique, I don't think that causes any problem for treeBreaker, but for other programs it might. Again to rule out possible problems, I would use unique names for each branch, maybe use non-numeric names like branch_1.

I also suggest that you run treeBreaker for at least 10 million iterations (maybe even 100 million) as you have a very big tree and the MCMC needs time to explore all the branches of the tree.

lindechun commented 6 years ago

Hi, Azim Firstly, I install treeBreaker on top of gsl that installed locally. and the test result of 'test data ' is not problems. Could you give me the binaries compiled on linux ? Thanks for the suggestion about -x, -y, -z (iterations) setting of run treeBreaker.

ansariazim commented 6 years ago

I don't have linux. But if you can run treeBreaker on testData then I would say it works fine. For the data that you sent me can you run it as follows treeBreaker -x 1000 -y 1000 -z 100 6829_ML.tree MDR output

and send me the output file.

lindechun commented 6 years ago

Hi, I ran treeBreaker on these files again,

and it worked fine after re-install treeBreaker. In addition, because the filename of output of treeBreaker is freewill, so I added the command-line arguments of treeBreaker.R

treeBreaker.R.zip