PICRUSt run fails to start.

Question

[Naren]

Dear all, I am trying to use PICRUSt for predicting metagenome of 16s profiles.
I have installed it successfully but when I run the very first script:

normalize_by_copy_number.py -i your_otu_table.biom -o normalized_otus.biom

It says:

Traceback (most recent call last):
  File "/usr/local/bin/normalize_by_copy_number.py", line 16, in <module>
    from biom.parse import parse_biom_table, parse_classic_table_to_rich_table
ImportError: cannot import name parse_classic_table_to_rich_table

What could be the reason and solution?

Answer 1

[5heikki]

$ python
>>>>import biom.parse
>>>>print biom.parse.__version__
>>>>quit()

What does it say?

PICRUSt site says biom (version 1.3.1)

You could always use the online galaxy version. Also, can you really trust the results you get from PICRUSt? You have a relatively short stretch of 16S rRNA that is e.g. 98% similar to the 16S rRNA of some reference genome. Based on this, the genome your 16S rRNA represents encodes the same functions as the reference genome? I don't think so. 16S rRNA genes drift apart far, far slower than the protein-coding gene contents of genomes. It might be somewhat decent strategy in like 20 years when we have a much more comprehensive set of reference genomes..

[Naren]

it says: 2.1.5 version ....

[5heikki]

Perhaps that function was removed/renamed. Google could maybe tell.

edit. https://groups.google.com/forum/#!topic/picrust-users/WzgQWDVYh9w