Possible bug in beta_significance.py?

[Lowry76]

Hi all,


I ran into a strange error when using beta_significance.py. When running it I got a

ValueError: No valid samples/environments found. Check whether tree tips match otus/taxa present in samples/environments error

which usually means there are differences in the naming of the OTUs in the biom file and the given *.tre. After checking I realized that make_phylogeny.py introduces single-quotes around the OTU names in the newick file which then leads to mismatches because beta_significance.py thinks they are part of the name. All other scripts, e.g. jackknife_beta_diversity.py, are fine with the quotes.

Question is: is it a bug in make_phylogeny.py or in beta_significance.py or is there something else I overlooked?

Cheers,
Lowry


PS: Experienced on MacQiime v1.9.1 as well as on linux Qiime v1.9.1

[Se Jin Song]

Hi Lowry,

Can you try running alpha_diversity.py using the otu table and the tree file you have, and letting me know if that works without errors?

Thanks,

Se Jin

[Lowry76]

Hi Se Jin,

Yes, alpha_diversity.py  worked with those files, too. To be more specific the problem is that in qiime's fast_unifrac.py in function delete_test() (in my case line 176) all branches are deleted because node.Name() returns the OTU names with quotes but the OTU table has them without, i.e., they don't match with the keys in the dicitonary env.

Cheers,
Lowry

[justink]

Hi folks. Certainly sounds like a bug, though I'm surprised I haven't heard of it before. Is there something unusual in your setup, like underscores in your sample names perhaps?

In any case, sounds like a bug. I'll file it and hopefully it'll be fixed soon. You seem to know quite a bit about python coding—if you're feeling really into it, go ahead and modify the code locally and see if you can fix the issue.

https://github.com/biocore/qiime/issues/2195