How do I interpret BLAST results

Question

[Alex]

Hi All,

I've used MAKER to produce some gene predictions on some of my largest contigs. I then subsequently saved the FASTA files it produced and used them to run a BLAST, with my query protein sequences being CYP genes found on NCBI. I am trying to analyse the results from this to determine whether I have any matches that are significant and indicate that I have some CYP genes in these contigs but I'm unsure how to determine, using the e-value, whether they are matches. Can anyone give me some idea how to analyse the BLAST results to determine this.

Thanks!

Answer 1

[Mensur Dlakic]

This is a topic with lots of resources that can be found by Googling.

What E-value threshold one takes depends on the exact intent. If you are looking to identify the exact match to a protein (an ortholog), you may need an E-value at 10^-40 - 10^-50, depending on protein and database sizes. A match with an E-value of 10^-20 is definitely a relative, but it may not perform the exact function as CYP. In some cases and with large databases, E-values all the way up to 10^-2 still indicate homology, although that would be of a distant variety.

To answer your question better, we need to know what exactly you are trying to prove, the number of proteins in the search database and E-values you are getting. Generally speaking, the more information you provide in your question, the better answers you will get.