I knocked down a protein and now looking for its substrates, so I need to be highly confident about my proteins ID. I used the Mass-spectrometer to analyse the samples and got my protein sequences data . After doing the comparison against the right database I got many of the 'unnamed proteins', so I blasted those unknown AA sequences against various sequence search tools, and did compare results in search for the closest homologous sequence .The main issue that I got totally different results,proteins, for blasting the same sequence on different DBs.
What criteria/parameter should I look at or rely on the most in order to be certain about my proteins ID? What is the most reliable factor to select the best hit match that most represents my sequence?
Your help would be highly appreciated.
Without seeing your data, I am just going to speculate. Usually, MS searches only give you peptide sequences along with the ID/name of the protein they think the peptides belong to. So in your case I suspect that some peptides were identified that either couldn't be confidently assigned to a protein or were assigned to a protein present in your database but that has no name associated with it. In the former case, trying other databases will probably not be of much help unless you get lucky. In the later case, you should try blasting the sequence from the unnamed protein against e.g. UniProt. One thing to look at first would be the confidence score associated with the peptides. I would discard all peptides below a certain threshold (which depends on the context, method used...) as there would be little guarantee that they are correctly identified.