Some question in Hmmer package

Question

[howenwy2]

I am currently using Hmmer package for sequence analysis (http://hmmer.org/). The problem is that I couldn't get any hit even if the querry data I used is from the same genome. Here is the detail of my problem:

I want to analyze the binding site in bacteria's genome. I aligned 4 sequence (13 nucleotide in each) with known binding site, and generate .hmm file by 'hmmbuild' command. After that, I used 'hmmsearch' command to find binding site in a bacteria genome (circular chromosome with 3805573 basepair).

Since the querry swquence is from the same genome, it supposed to have at least 4 hits in the result; however, no hit could be detect in my result.

I don't know how do I do to fix this problem? It seems that I need to change some default value.

I use OSX 10.10.5, and the package version of hmm is the latest one (3.2.1), could that fit? If that couldn't fit my system, what version will you recommend?

[Asaf]

It doesn't have to find the original sequences. Does the resulting motif has information in it? You can have an idea by plotting the MSA of the input. I suspect that four sequences is too little to generate a meaningful motif.

[howenwy2]

Thank you so much for the reply. I am new to bioinformatics, I am sorry that I am not sure if I understand your reply totally. The following is my MSA of the input:

CLUSTAL O(1.2.4) multiple sequence alignment

1. gene1 TTTGAGTGTGTTA 13
2. gene2 TTTGATCTGGTTA 13
3. gene3 ATTGAGGTAGTTA 13
4. gene4 TTTGAGGCTATTG 13

I need to find the (T/A)TTGANNNNNTT(G/A) in my genome sequence, and gene1 to gene4 is also the sequence from the same genome, and I need to find that sequence from the other genes in this genome. Could that be possible that the sequence is too short or too little so that I couldn't find that in the genome?

[Jean-Karim Heriche]

It seems that you're looking for a well defined motif. In this case, you could just search using regular expressions, e.g. /[T|A]TTGA.{5}TT[G|A]/g assuming the occurrences are non overlapping in which case, you probably can use look arounds like /(?= [T|A]TTGA.{5}TT[G|A])/g.

[Joe]

HMMs are probably a bit overkill for this. Since you already know the degenerate sequence, try using a fuzzy matching aligned like fuzznuc from the EMBOSS package.