Co-occurrence analysis

Other Bioinformatics Tools
1

Hello Everyone,
I am interested to find an interaction or relation between fungal and bacterial abundance like attached figure "co-occurrence%20analysis"
using two sets of data (otus.biom from 16S and 18S rRNA). Please, anyone can suggest me which kind of package (version) or analysis should I follow? If any package with support of Qiime?
I found an analysis named Co-occurrence but could not get proper package to do analysis?

Thanks,
Sincerely,
Khemlal

1 Like
2

Hi @khemlalnirmalkar,
I guess the best available tool I know is Conet http://psbweb05.psb.ugent.be/conet/index.php
http://journals.plos.org/ploscompbiol/article?id=10.1371/journal.pcbi.1002606
Conet is a plugin of cytoscape http://www.cytoscape.org/. this means you have to install Cytoscape first. Both have plenty of tutorials and are relatively easy to use. You can find a nice tutorial for conet here http://psbweb05.psb.ugent.be/conet/microbialnetworks/conet_new.php.
A nice feature of conet that was added in the recent update, is using biom files directly without the need to convert or transform your data. If you follow the tutorial in the link above you will find your way easily and you can always ask cytoscape-helpdesk https://groups.google.com/forum/#!forum/cytoscape-helpdesk

Good luck and I hope this answers your question

Ahmed

1 Like
3

Hello @Ahmed_Abdelfattah,
Thank you so much for your suggestion. These links are really useful for me.

I tried CoNet and Cytoscape with sample files and it ran successfully but when I tried with my files, shown some errors which is in below. Can suggest me what could be the possible mistake in my files.
I tried to make my files in similar manner as this tools have in sample files.

Error message:
java.lang.IllegalArgumentException: After filtering, less than 2 rows remain in the input matrix!

Sincerely,
Khemlal

1 Like
4

Hi @khemlalnirmalkar

I’m glad to know you were able to get conet to work!
This error message, and I’m quoting, “means that your input matrix contains less than two taxa, so no associations can be computed. Unless you provided an input matrix with only one taxon row, the reason is that after filtering, most taxon rows were removed. You can use less stringent filtering settings, but at the risk of increasing false positive edges. To change filter parameters, please open the Preprocessing and Filter menu and alter the parameters in the upper left Input filtering box.”
Please refer to conet documentation and read the FAQ http://psbweb05.psb.ugent.be/conet/faq.php. You will find most of the errors that you will encounter in the FAQ link

Good luck,
Ahmed

2 Likes
5

Hi Ahmed,
Once again thanks for your suggestion.
But I am sorry, I could not understand properly what is the meaning of this error message.

Please can you explain me little bit more.
Here I attached my file (text file from biom, generated from qiime1 using silva104 database18S_new_otu_table.txt (162.9 KB)
), if you have time can you look at once and tell me what is the mistake?

Kindly ignore the second attachment 18S_new2_otu_table.txt file.

Sincerely,
Khemlal
18S_new2_otu_table.txt (157.4 KB)

6

This topic was automatically closed 31 days after the last reply. New replies are no longer allowed.