Entering edit mode
8.4 years ago
Neha shri
▴
30
Hi everyone,
I am trying to write a program to screen protein sequences based on their amino acid substitution score.
Initially, I tried to do it by simply placing a score of 1 (at relevant substitutions) and 0 where the substitution is deleterious. Is the logic evolutionarily meaningful?
Another thing is, can I use blosum substitution scores in place of the above mentioned method? I know there is BLAST (ppos available for this). Just want to do in a different way.
Any pointers will be appreciated.
Do you really need to re-implement this, e.g. to practice? If not you could use any local or global aligner, e.g. ssearch, ggsearch. It has various substitution matrices built in, including blosum50.
Thank you :)
I didn't understand what you were trying to do. Perhaps you can use global alignment (with BLOSUM62), this will give you a measure of how much two proteins are similar. You can use needle of EMBOSS for instance (there are several publicly available EMBOSS servers and you can download it)
Initially , tried to do it by simply placing a score of 1 (at relevant substitutions) and 0 where the substitution is deleterious. Is the logic evolutionarily meaningful?
How do you know that the AA substitution is deleterious? Did you mean nucleotide substitution leading to stop-codon?
I meant , lets say a hydrophobic non polar amino acid residue in the core of a protein is replaced (substituted) by a polar one which is meant to be on the surface. That can pretty much reduce the stability .
That's arbitrary logic. You're better off implementing BLOSUM (which you are doing, and is the right way to go).
Thank you :)
Also, it's BLOSUM (from BLOcks SUbstitution Matrix), not blossum :)