Hi Arnaud, Let me explore this further, we will get back to you on this shortly.
Cheers Syed > Hi Syed, > > Yes, we do replicate it, although this particular case occurs when the > translation table is 11, so not for vertebrates. > > The rule is that if the start codon is one of the valid ones according > to the translation table used, usually 11, see > http://www.ebi.ac.uk/embl/Documentation/FT_definitions/feature_table.html#7.5.5 > > the codon codes for a methionine. For translation table 11, these are > TTA. CTG. ATT, ATC. ATA. ATG and GTG. > > To me, it is not a post-translational modifications and should not be > represented in the database by any extra attributes. > It is just a matter of fact that the translation code should be aware > that when the CDS is complete, the first codon is always a methionine > even if the genetic code says it is a valine or a leucine. > As mentioned in the previous email, the Ensembl API doesn't need any > extra information in the schema, and can deal correctly with such cases. > > cheers, > Arnaud > > Syed Haider wrote: >> >> Hi Arnaud, >> >> As I understand, you are replicating 'ensembl vertebrates' data model >> for your 'bacterial genomes'. In Ensembl Vertebrates, the information >> for changing the first amino acid to 'methionine' comes as a >> post-translation modification, which sits in a separate database >> column. Please see if thats the case in your mart too. >> >> Thanks >> Syed >> >> >> Arek Kasprzyk wrote: >>> >>> On 13/11/08 1:00 PM, "Arnaud Kerhornou" <[EMAIL PROTECTED]> wrote: >>> >>> >>>> Hi everyone, >>>> >>>> We are setting up a Mart database for bacterial genomes. I was looking >>>> at the peptide sequence attribute. So far it seems fine except >>>> sometimes >>>> the first amino acid which should be a Methionine is displayed as a >>>> different amino acid. >>>> >>>> Roughly, the rule is, if the CDS is complete and if the first amino >>>> acid >>>> is not a methionine, then replace it by a methionine: >>>> >>>> e.g. see Bioperl code, Bio::PrimarySeqI->translate(): >>>> >>>> ## Only if we are expecting to translate a complete coding region >>>> if ($complete) { >>>> # ... >>>> # if the initiator codon is not ATG, the amino acid needs >>>> to be >>>> changed to M >>>> if ( substr($output,0,1) ne 'M' ) { >>>> if ($codonTable->is_start_codon(substr($seq, 0, 3)) ) { >>>> $output = 'M'. substr($output,1); >>>> } elsif ($throw) { >>>> $self->throw("Seq [$id]: Not using a valid initiator >>>> codon!"); >>>> } else { >>>> $self->warn("Seq [$id]: Not using a valid initiator >>>> codon!"); >>>> } >>>> } >>>> } >>>> >>>> How do you translate CDS sequences in Martview, is it made on the fly >>>> within the Biomart perl API ? >>>> >>> >>> Hi Arnaud, >>> Yes it is. Let us investigate what's going on there and we'll get >>> back to >>> you with a suggestion >>> >>> A. >>> >>> >>> >>>> Is there a way of fixing this translation issue ? >>>> >>>> Thanks >>>> Arnaud >>>> >> > ====================================== Syed Haider. EMBL-European Bioinformatics Institute Wellcome Trust Genome Campus, Hinxton, Cambridge CB10 1SD, UK. ======================================
