Increasing the percent identity of similar structures/sequences retrieved using blast.pdb() and hmmer()

[Former user]

I am not sure if I have completely missed something obvious, but I have been trying to find a way to increase the search parameters when using either the blast.pdb() or the hmmer() functions. The sequence I am using only returns a few sequences however I am looking to try and retrieve more than that, with a range of percent identities. Have I overlooked the obvious or is this not an implemented function? Thanks!

[Xinqiu Yao]

Hi,

‌

Can you give an example for what you did along with some output? We then may reproduce the problem and find out what was going on.

[protein_fashion]

Thanks for getting back to me. It’s not a problem per say, but an inquiry that I may be missing the point of. When I use blast.pdb() or hmmer() for a pdb file in order to retrieve similar sequences, I only get about 9 back. When manually searching on the blastp website, I get more hits by allowing a wider percent identity. I have tried looking for a way to change the search parameters for the two functions but haven’t found any. For instance:

hmmer(get.seq("4owx_B"))

Will return 28 hits, but all of them are of a high percent identity to the input sequence, and I am looking to get a higher number of hits to investigate and compare, from more than just those that occur in that search. To do so I need to change the percent identity cutoff, so I can retrieve more results. Is there a way to do that using Bio3D and I have just missed it, or should I manually import a set of filtered sequences? Thanks!

[Xinqiu Yao]

Bio3D uses default parameters to run BLAST or HMMER and returns whatever hits obtained by the servers with those parameters. So far, there is no options to directly restrict results by a threshold of sequence identity or significance score. What you might need to check is that you have chosen the same database in Bio3D and website. Note that Bio3D uses “pdb” by default whereas NCBI uses “nr”. This may cause the difference.

[protein_fashion]

Awesome thanks so much Xinqiu, I had a suspicion that was the case, so I will work from there. Thanks for the help.