I need to know if the primary sequence provided by this database is trustworthy. I'm working on a transmembrane protein for which no complete x-ray structure is available. The point is that recently the x-ray structure of one of the members this protein family has been discovered. When we do the sequence alignment of this protein (the seq obtained from UniProt) and the new structure, it seems two residues are not defined in the seq of my protein. I don't know whether I have to trust the UniProt seq or I have to model those 2 residues for my protein.
In the same way NCBI does, the Uniprot group has two different sequence databases, the first one is called UniProtKB/TrEMBL and is populated with unreviewed, automatically annotated sequences, the second one is called UniProtKB/Swiss-Prot and is used for reviewed and manually annotated sequences.
Nevertheless keep in mind that if your protein and the one you have downloaded are from different organisms its OK to see some degree of divergence in the primary sequence. This divergence can vary and its kinda hard to say how much, but some functional and structural domains accept a certain degree of variation in its primary sequence.
Personally, I would BLAST the UniProt seq and your candidate sequence and see, which fit is better. Preferably, as the DB for BLAST - choose cDNA to avoid mRNA rearangements.
Hppe that helps,
- mno
More than one sequence are available (i.e. mutated and/or alternative exons arrangement), so follow the Marcin's suggestion.
Furthermore, it happens that the X-ray structure derive from sequence that have been further mutated in order to obtain a more stable crystal (e.g. ad-hoc addition of a salt bridge). Check also the reference article related to your pdb file.
In the same way NCBI does, the Uniprot group has two different sequence databases, the first one is called UniProtKB/TrEMBL and is populated with unreviewed, automatically annotated sequences, the second one is called UniProtKB/Swiss-Prot and is used for reviewed and manually annotated sequences.
Nevertheless keep in mind that if your protein and the one you have downloaded are from different organisms its OK to see some degree of divergence in the primary sequence. This divergence can vary and its kinda hard to say how much, but some functional and structural domains accept a certain degree of variation in its primary sequence.
I agree with Felipe on the twin nature of UniProt about annotation (TrEMBL/Swiss-Prot). But, is it correct that in Uniprot only protein sequenced (not DNA/RNA sequenced) data are deposited?
According to the UniProt; "about 98 % of the protein sequences provided by UniProtKB are derived from the translation of the coding sequences (CDS) which have been submitted to the public nucleic acid databases, the EMBL-Bank/GenBank/DDBJ databases "
Dear Zeynab,
I suggest that you contact the UniProt helpdesk ([email protected]), specifying the UniProt accession number and PDB identifier. A curator can then look into your specific example, add the PDB cross-reference and update the UniProtKB entry where applicable.
Hey Zeynab,
For the two residues that did not match with the uniprot sequence, were these residues by chance alternate residues?
I recall sometimes that you might get alternate residues at the same point in the pdb file as the resolution at that co-ordinate was not resolute enough to unambiguously give you a single residue.
- Badges
- Science topic
- Similar topics
- Bioinformatics
- Bioinformatics and Computational Biology