Hello /sci/. I am working on a program which I intend to use for genome analysis.
In part of this program I would like to use a hidden markov model to search a whole genome for a specific gene.
I am using HMMER to build an HMM from a multiple sequence alignment of several homologs to the gene I'm looking for.
My issue is that I cannot find a way to use the HMM I created to scan through a genome like you would with a sequence alignment.
Most of the HMMER methods, like hmmsearch, search through a database for matches but from my own experimentation it seems like I can't just use my genome as a database.
It seems like to make a database for this, I would have to cut up my genome into all possible chunks the same length as the gene sequence I'm looking for, and then hmmsearch through that database, but that seems far to computationally intensive, and would also give duplicate hits for the same DNA region.
A paper I was reading used a function called 'hmmls', but this is no longer in the latest version of HMMER.
They also didn't go into any detail at all on how they did the searching using said function.
Do any of you guys have suggestions?
>inb4 use BLAST
I want to pick up on somewhat more distantly related homologs which blast would not pick up
>inb4 use PSI/DELTA BLAST
I'm not sure how I would use the PSSM generated to search through my genome as opposed to a database, so I'd rather stick to my current method unless one of you knows how to do this.
In part of this program I would like to use a hidden markov model to search a whole genome for a specific gene.
I am using HMMER to build an HMM from a multiple sequence alignment of several homologs to the gene I'm looking for.
My issue is that I cannot find a way to use the HMM I created to scan through a genome like you would with a sequence alignment.
Most of the HMMER methods, like hmmsearch, search through a database for matches but from my own experimentation it seems like I can't just use my genome as a database.
It seems like to make a database for this, I would have to cut up my genome into all possible chunks the same length as the gene sequence I'm looking for, and then hmmsearch through that database, but that seems far to computationally intensive, and would also give duplicate hits for the same DNA region.
A paper I was reading used a function called 'hmmls', but this is no longer in the latest version of HMMER.
They also didn't go into any detail at all on how they did the searching using said function.
Do any of you guys have suggestions?
>inb4 use BLAST
I want to pick up on somewhat more distantly related homologs which blast would not pick up
>inb4 use PSI/DELTA BLAST
I'm not sure how I would use the PSSM generated to search through my genome as opposed to a database, so I'd rather stick to my current method unless one of you knows how to do this.
