Submitting Aa Sequence

Question

[Michael]

I am working with 454 sequences obtained by a different research group. According to their sequence annotations, we have identified the members of a family in which we are interested. By computing, we have obtained the putative ORF and translated to aminoacid sequence in the right frame.

Is it possible to submit the protein sequence to a public DataBase (UniProt, for example), although the original nucleotide sequence is already published and obteined by a different group?

Answer 1

[Larry_Parnell]

Generally, no, you would not submit the amino acid sequence because you did not sequence a protein or peptide. You sequenced nucleic acid and so that is what you submit. Sequin from NCBI is a good tool for this purpose.

You can certainly provide a conceptual translation of the submitted sequence and in so doing there will be a protein sequence in the appropriate databases. In fact, when you submit this way - a DNA sequence - you will identify the ATG start codon, the termination, splice sites/exon sizes (if any), as well as that protein sequence. All this other information is very valuable.

[Michael]

Thanks Larry. I've written UNIPROT too. Here their answer : What you describe sounds like a case for a 'third party annotation' (TPA) submission to EMBL-Bank (or GenBank): Submit your computationally analyzed coding sequences to EMBL-Bank/GenBank and the translations will be automatically imported into UniProtKB-TrEMBL. You cannot submit the translations directly to UniProtKB, however.

Here is the link to the EMBL submission info:http://www.ebi.ac.uk/embl/Documentation/information_for_submitters.html

[Larry_Parnell]

Yeah, just what I thought - but I'd never have arrived at such legal-sounding language.