Align a position weight matrix against a sequence by using bioperl

Question

[francisco.abad]

I everyone.

I'm trying to find a certain probabilistic based pattern in a DNA sequence by the using of position weight matrix. That is to say, I specify several nucleotides with their probabilities for the same position.

I'm using perl and bioperl modules to achieve this. I was able to create an psm object using the modules Bio::Matrix::PSM::SiteMatrix and Bio::Matrix::PSM::IO, but I have no idea about how to perform an alignment against a sequence using this matrix.

What module could I use to perform the alignment?

Thanks in advance!

Answer 1

[Joseph Hughes]

You probably want to use Bio::Tools::dpAlign

[francisco.abad]

Hi Joseph, thanks for your answer. I've checked your link, but I think it is used to perform alignments between sequences. In my case, I need to find matches of a position weight matrix in a sequence. A position weight matrix is a matrix that have the following structure:

        A    C   T   G
----------------------------------
POS1|   50  20  20  10
POS2|   20  20  10  50
POS3|   50   0  50   0
POS4|   25  25  25  25

This kind of matrix represents the probability of each nucleotide for each position. The idea is to find matches in a sequence that were accomplished by the matrix.

This feature is useful if you want to find protein binding sites in the genome, for example. I've seen applications that can perform this task, but I want to do it in a programmatically way.

Answer 2

[francisco.abad]

Hi, finally I found a web service that performs this task at Regulatory Sequence Analysis Tools (RSAT) webpage. Concretely I'm interested in matrix-scan tool, but you can find another interesting tools too.

The web browser version of scan-matrix is at http://tagc.univ-mrs.fr/rsa-tools/matrix-scan_form.cgi. The documentation about how to use it in a programmatically way by using SOAP web services is at http://tagc.univ-mrs.fr/rsa-tools/.

Cheers, Fran.