reciprocal BLAST meaning

Question

[anuj2054]

I am Anuj, a PhD student, from the University of Oklahoma. I was trying to replicate your PNAS paper titled : " Major diversification of voltage-gated K+ channels occurred in ancestral parahoxozoans" I wanted to ask a couple of questions about it and maybe you cold point me to resources or other people. My question is about the first paragraph in the methods sections which is

Cnidarian and Mnemiopsis Shaker and KCNQ family K+ channel genes described in this study were identified and compiled through comprehensive BLAST (46) searches of genome drafts, transcriptomes, and gene predictions of Mnemiopsis leidyi (41) andNematostella vectensis (73), and for KCNQ only, Hydra magnipapillata (66), and Acropora digitifera (74) andOrbicella faveolata. Multiple bilaterian members of each channel type were used as query sequences, and reciprocal BLAST searches of identified sequences against bilaterian databases were used to classify the sequences before phylogenetic analysis. Most queries identified all voltage-gated K+, Na+, and Ca2+channels, but reciprocal searches sorted target sequences by gene family and were used to refine gene predictions when necessary.

I have found the wording for this paragraph really confusing. Was protein prediction first performed on the Mnemiopsis ,Nematostella, Hydra, Acropora, and Orbicella transcriptomes and then the Shaker, KCNQ genes used as a query ? Or Was the identified proteins from the transcriptomes used as a query for the Shaker and KCNQ genes ? After the initial BLAST another BLAST was performed of the results as query against the bilateran ion channels as database ?

I would greatly appreciate it if you could elucidate this further for me.

Answer 1

[Ram]

I'm guessing you copy-pasted an email here, and hence the phrasing that you've used. I'd edit that if I were you.

In my experience, reciprocal BLAST refers to when you have two datasets, and you each as the database for the other to be queried against. It generally gives you mutually-best matches. My group's understanding when I did this was: If a sequence A1 in set A aligns best to sequence B5 in set B, but B5 aligns to some other sequence in A significantly more so than it does to A1, you know A1 had to be modified (with gaps and mismatches) a lot for B5 to be the best match in the first BLAST. A mutual top level match indicates a high chance for evolutionary homology if the datasets are from related species.