[Bioperl-l] (no subject)

[Dhr. R. de Jonge]

Dear all,

I've recently started working with the bioperl package. However, as I
wasn't trained a bioinformatician (rather molecular biologist) I'm now
and then facing difficulties starting up. Though, using some books
from James Tisdall and documenting the bioperl manuals I figured out
much of the stuff.

Still I'm facing a problem I don't seem to get fixed easily. The
problem is as follows:

1) I've set-up a blastable database of cDNA/transcript sequences of a
complete set of fungal genomes (and produced a protein blastable
database from this one).
2) I've checked this database for a particular HMM domain using the
Bio::Tools::Run::Hmmer module and some custom scripting. Using this I
generated a file containing a simple fasta header and corresponding
domain-sequence (protein).
3) However, from an evolutionary view I want to look at non-synonymous
versus synonymous mutations (using the transcript data I started
with). To perform this test I'd like to use the bioperl PAML module.
Naturally; I need the 'original' cDNA sequence to test my hypotheses.
(the question then is: how to obtain this cDNA sequence?).

I tried to solve this issue by blasting the obtained protein sequence
to the cDNA database using the tblastn package. I specified in a
script the number of hits to be screened (i.e. 1), and used the
trunc() method (using the blast hit sequence) to get the corresponding
cDNA sequence. However, it seems not all hits are perfect matches,
i.e. on some occasions not the original (corresponding) sequence
object is retrieved, leaving me with false cDNA --> protein lists.

Coming to the actual question: Is it possible to circumvent this whole
(sort of false) procedure and use e.g. the Bio::Annotation modules? If
yes, can you give some hints/feedback on setting-up a method for this?
I'd like to hear your thoughts on this one.

Many thanks in advance.

With regards,

Ronnie de Jonge

MSc. Ronnie de Jonge
Wageningen University
Department of Plant Sciences, Laboratory of Phytopathology
Binnenhaven 5, 6709 PD  Wageningen
Tel.: +31-317-483224