I was wondering if predicted protein sequences, which sometimes come out of the 'Gnomon annotation program' of NCBI, can be used for multiple sequence alignment in common tools such as Clustal omega? The ultimate objective is to identify the conserved domains and then design the primers for PCR. My gene of interest is not sequenced in the species in question. The whole genome shotgun sequences are available for that species but not annotated yet. The protein sequences of this gene are available from other species but mostly 'predicted'.
Thanks in advance for taking time out for replying.
Those predicted sequences can be perfectly good, and other times they are not. Can you find the same protein in related species? As a quick check, you can create an alignment of the orthologous protein sequences and view in an alignment viewer (like jalview or clustal X). See if the sequences most agree with one another. If your predicted sequence looks drastically different than the same protein in other species, then the prediction might not be of the best quality.
If your predicted protein sequence doesn't seem to be of good quality, you can use those related protein sequences as a template to do a prediction of your own in the genome of the target species using the exonerate program.
http://www.ebi.ac.uk/about/vertebrate-genomics/software/exonerate
- Badges
- Science method