OMA current state and its updates

Question

[natasha.sernova]

Dear OMA team!

I like your program, since there are very few tools that help to find “true”-orthologs, and your papers convinced me OMA

is one of them.

I have several questions to you.

1) I don’t like to run OMA on our cluster, it takes a lot of time, uses many nodes and my colleagues do not approve it. It turned out now it is possible to run it locally – I need to find orthologs to several proteins only. It works but BLAST finds more protein hits… Some of them are not “true”-orthologs, but many are.

2) What is your default database? I found some proteins have not been there yet. I need to run OMA in many bacteria for a single protein, or even in all finished bacterial genomes for a single protein. Is it possible right now with OMA ?

3) What would be your advice -what script should I use to make OMA run as fast as possible? Our cluster is rather specific – scripts suggested in your papers do not work on it.

4) What is the latest update – I have not found any articles published later than in 2015. How can I update my old OMA version?

5) Where can I find a detailed description of the current OMA algorithm?

Thank you very much indeed! Sincerely yours, Natalia

Answer 1

[Christophe Dessimoz]

Hi Natasha,

1. Smith-Waterman alignments are necessarily more sensitive than blast alignments. The reason you might get more hits is likely due to differences in length and minimum score thresholds. You can play with these parameters in OMA Standalone (but as you hint, lowering false negatives will increase false positives)
2. By "default", the databases are those exported from OMA (http://omabrowser.org/export). About 100 genomes get added or updated at each release. You can get the details here: http://omabrowser.org/cgi-bin/gateway.pl?f=InfoMatrix
3. Export genomes from OMA to avoid computing some of the all-against-all, use parallelism. We are working on some speed improvements which will hopefully be made available sometime in 2017.
4. We plan to publish within a few months a paper outlining the most improvements on the OMA algorithm. Stay tuned! Meanwhile the improvements are already available in the latest version of the OMA Standalone software (http://omabrowser.org/standalone)
5. Again stay tuned (but if you can't wait send me an email and I will gladly share with you our latest draft!)

Best wishes, C.

[Christophe Dessimoz]

The updated OMA algorithm paper is now available: http://doi.org/10.1093/bioinformatics/btx229

[natasha.sernova]

Dear Christophe, THOUSAND THANKS!!!

[natasha.sernova]

Dear Christophe, Many-many thanks! If I still have some problems with OMA, I'll send you a message.