Translate ESTs using related proteome

[k-gun12]

I've got denovo assembled RNAseq EST contigs from two subspecies of insect very closely related to one which has a nicely curated proteome. I want to compare protein sequences, but I need to first translate the nucleotide EST contigs. I can use a six frame translation and find the orthologs and/or grab ORFs, but those are open to assembly errors and/or indels throwing off the coding frame. Does anyone know of a way to grab the "most likely" peptide sequence of my ESTs using a reference proteome or nucleotide sequence of a conspecific?

Thanks!

[krobison]

I'm afraid I don't have a ready-made tool for this, but the usual approach I use is to identify the best match with BLASTX and then use the venerable ESTWise (with the -pep option) to generate an easily parsable file which your favorite text mangling language can pop the translation from.

If you have a cluster, then spawning the ESTWise nodes using a queue manager such as Grid Engine is highly advised; the ESTWise jobs can take a while. You can also restrict the strand based on the BLASTX data; this should roughly halve your job.

In theory there is a Perl wrapper around ESTWise, but I've never had luck with it.