There is a published substitution matrix for intrinsically disordered proteins that I would like to use for a BLAST search, but I am unable to find a program that supports uploading a custom matrix. I prefer to use R for my computational biology, but I will use what ever is needed to support the matrix. Any recommendations or tips? Thanks!
Although old, yes, there is. Legacy Blast 2.2.12 (or 2.2.13, if I well remember) accept custom matrices. I've used this feature successfully in 2010. The following versions of Legacy Blast and (I think but not 100% sure) in all Blast+ series the matrices are included in the program, so you cannot change them. The fact that the version that accept matrices is old means that it is probably less efficient computationally, but if you can run it in your machine, you'll get good results as blast heuristic is essentially the same since 1997.
Part of the arguments to hardcode the matrices, at the time, were: 1. it made the blast search more efficient; 2. most of the researchers use only the known matrices. and; 3. no reviewer would "believe" in a newcomer without all the testing that current matrices have already passed.
I think this was an unnecessary added stiffness, as new ideas in sequence analysis may translate into new substitution matrices. This way, NCBI blast team saved themselves from hard time of being blamed by someone whose substitution matrix didn't "work well".
I think you can still download older versions of blast and include your own matrices to find new stuff. That being said, don't forget, Blast matrices are trash in, trash out.
- Badges
- Science topic
- Similar topics
- Mathematics
- Statistics
- Statistical Software
- R Statistical Package