Reciprocal Best Hits across more than 2 species

Question

[mforthman]

I have exons from a reference taxon. I want to now identify putative homologs from transcriptomes of 5 other species using the reference exons and reciprocal best hits BLAST. I've seen a number of threads and scripts that primarily focus on performing RBH with two species only, and I realize this is what RBH does in the background. However, I'd like to do this across all species so I know not only what is a putative homolog between species 1 and reference species, but between species 2–5.

The idea is that I can then extract exons that are identified across all species, or alternativley, that are found in at least two species to create a variety of datasets. Is anyone aware of any scripts that can do what I'm needing done?

Answer 1

[natasha.sernova]

You need to find orthologs, am I right?

The approach you have chosen is not the best one, see the post below.

BLAST Reciprocal Best Hit question NCBI

Reciprocal best hits are not a logically sufficient condition for orthology

https://arxiv.org/pdf/0706.0117.pdf

There are a couple of rather recent papers below:

Quickly Finding Orthologs as Reciprocal Best Hits with BLAT, LAST, and UBLAST: How Much Do We Miss?

http://journals.plos.org/plosone/article?id=10.1371/journal.pone.0101850

Bidirectional Best Hits Miss Many Orthologs in Duplication-Rich Clades such as Plants and Animals

http://www.ncbi.nlm.nih.gov/pmc/articles/PMC3814191/

There is a very detailed discussion in this post, read it.

What Is The Best Method To Find Orthologous Genes Of A Species?