User Defined Pssm For Clustal
1
2
Entering edit mode
12.3 years ago
jwillis0720 ▴ 20

Hi,

I'm not exactly sure what tool to use, but I want to make a phylo tree using my own PSSM I generated off some structural constraints. Anyway, clustal says it has the option to use user defined matrices rather than the hard-coded ones. It says this needs to be in blast format. I was wondering if anyone could tell me the easiest way to get my pssm into "blast" format. I have tried downloading the ASN.1 files and they are pretty convuluted. Is there any documentation explaining on how these matrices should go?

I have tried to generate some psiblast output using

psiblast -db somedatabse -in_msa input.fasta -out_ascii_pssm pssm.txt

However, when I try and put this pssm back into clustal it says "too many entries"

I'm not dead set on using clustal, I just want to use my matrix. Does anyone have any suggestions?

clustalw pssm • 3.6k views
ADD COMMENT
1
Entering edit mode
12.1 years ago
Hamish ★ 3.3k

ClustalW can use a user defined scoring matrix (referred to as weight matrix in the help) in the "slow" pairwise alignment method for generating the guide tree and in the multiple sequence alignment phase of the progressive alignment process. This scoring matrix is expected to be in the same format as is used by NCBI BLAST (and many other programs), see ftp://ftp.ncbi.nih.gov/blast/matrices/ for examples. Please note that for protein alignments ClustalW's default behaviour is to use a series of matrices in order to optimise the scoring, when a user supplied matrix is specified only the single matrix provided is used.

A Position-Specific Scoring Matrix (PSSM) is different from the scoring matrix used by ClustalW, in that it describes the frequency of each amino-acid at each residue position, instead of a model of how likely one amino-acid is to be replaced by another. You can use the NCBI's PSSM Viewer to examine a PSSM and see how they are constructed. A PSSM provides a mechanism for providing scoring that is context sensitive.

If your PSSM was generated from a sequence alignment (or something that can be converted into a sequence alignment) then you could use the alignment to generate a phylogentic tree. For instance an NCBI BLAST or PSI-BLAST result can be used to generate an alignment using tools such as DbClustal or MView, which will preserve the aligned segments appearing in the alignment. The resulting alignment could then be used as input to a phylogeny tool.

As far as I can see there is no way to use a PSSM to directly derive a tree, so your best bet is to get an alignment somehow either based on the source data used to generate the PSSM or by using the PSSM to perform a sequence search.

ADD COMMENT

Login before adding your answer.

Traffic: 2939 users visited in the last hour
Help About
FAQ
Access RSS
API
Stats

Use of this site constitutes acceptance of our User Agreement and Privacy Policy.

Powered by the version 2.3.6