Is A "Gene-Like," Returned By Ncbi Blast, Indicative Of Gene Family Membership?

Question

[Andrea]

I'm looking to sequence HSD in a non-model organism. I've been using degenerate primers for awhile and finally got a successful sequence. BUT when I enter it into BLAST, I get high sequence identity with HSD-like sequences in species closely related to mine (and not). What does this "like" mean? Is the gene sequence "like" that of HSD in other species? Does it say anything about the protein homology/function? Should I not even bother with this sequence? Any help would be greatly appreciated! Thanks!

Answer 1

[Larry_Parnell]

More details would help in assessing what you have and how relevant it is. For example, what is HSD? How long was the query sequence for BLAST? Which BLAST algorithm did you use and how long/how much sequence identity did you see in the results?

There can be a small or big difference between gene and gene-like. In the extreme case, the -like turns out to be a pseudogene. On the other hand, the -like designation may simply indicate that the function is implied because the "-like" gene is very similar to another gene with that proven function. This uncertainty can be cleared up with functional tests.

In short, I would not toss the sequence because it is a gene-like and a "pure" gene. You're coming at this from a non-model organism standpoint and so your gene(s) of interest may be most similar to -likes than to other genes in the family. Take your sequence, perform a multiple sequence alignment (trimmed to the regions you have as sequence data, or full-length gene if you have that), and then do some phylogenetic analysis to get an idea if your sequence is way out there (i.e, nonsensical) or fits well within this gene family. Repeat this with another gene from the same organism in order to compare.

[Andrea]

Thank you for your response. The gene is actually 11 beta hydroxysteroid dehydrogenase- a glucocorticoid metabolizer and I'm looking in a wild bird. I've cleaned up the sequence I have well and when I align it, I get 11BHSD-like in chicken and well as human/mice (but not as closely related obviously). There is no 11BHSD in chickens, but the sequence I have does not have great homology with mouse 11BHSD. As I'm not working with a model organism, knock out of any kind is out of the question and most other genes in this family were generated by me.

[Larry_Parnell]

No 11BHSD in chicken tells me that the 11BHSD-like in chicken is an HSD but one with either unknown substrate or unknown product or both. That's OK - no one has done the enzymatic reaction yet (except the chicken!). Your sequence still sounds fine to me. I would still tell you to perform the phylogenetic analysis.