Reader Comments

Post a new comment on this article

DNA matches to novel Rec genes?

Posted by KForslund on 04 Apr 2011 at 23:37 GMT

When reading this interesting work, a question occurs to me, which should be relatively easily testable. I suspect you might already have tested it, so I´ll ask it here.

Have you attempted to build DNA-based profiles of the "unknown" clusters and search the nucleotide, rather than the protein, databases for them? That is, could it be that the sequences represent some kind of degraded
pseudogenes, so that while there are no protein-coding versions present in sequenced genomes, there may be nucleotide motifs matching them?

No competing interests declared.

RE: DNA matches to novel Rec genes?

jeisen replied to KForslund on 04 Apr 2011 at 23:52 GMT

That is an excellent question and idea and we have not looked at that directly. We did not really do any analysis at the DNA level per se. So for example, we did not look at codon usage or nucleotide composition which could have been useful to try and identify the origin of the sequences or predict expression levels. in terms of whether these are pseudogenes, this seems unlikely because the sequences seemed to have roughly the same level of similarity to known RecAs throughout their length. if they were pseudogenes we might expect themm to lose similarity after some point (if there was a frameshift). But we did not think of doing the analysis you suggest and really should ... Great idea

Competing interests declared: I am the senior author of the paper