[EMBOSS] Custom Scoring Matrix

Allen Smith easmith at beatrice.rutgers.edu
Fri Dec 2 21:27:47 UTC 2005


In message <072b01c5f781$d32498c0$5022db82 at GOLHARMOBILE1> (on 2 December
2005 15:49:01 -0500), golharam at umdnj.edu (Ryan Golhar) wrote:
>I've created a custom scoring matrix of log-odds probabilities for
>comparing certain types of sequences.
>
>I'm not trying to use stretcher to compare two sequences using my
>matrix.  I'm passing in the matrix as an option to stretcher using
>-datafile:
>
>stretcher -asequence a.fasta -bsequence b.fasta -datafile ./matrix.mat 
>
>I get an error "EMBOSS An error in ajarr.c at line 1752: Bad integer
>conversion".
>
>My matrix looks something like this:
>
>        A       C       G       T
>A       1.1619  -0.9808 -0.7169 -1.2463
>C       -1.1165 1.54077 -0.8689 -0.8288
>G       -0.7911 -0.8621 1.5170  -1.1425
>T       -1.2419 -0.7391 -1.0441 1.0995
>
>I take it from the error, strether requires integers.

It appears so - it is using the AjPMatrix, not AjPMatrixf, datatype, for
instance.

>Is that true, or is the error generic?  If it requires integers, then I
>will need to find a way to normalize this matrix to integers.  Has anyone
>done this before?

Yes. If you're needing all the decimal places, I would suggest simply
multiplying by 10000 (and ditto for gap open/extend penalties). That at
least seems to work...

	    -Allen

-- 
Allen Smith                       http://cesario.rutgers.edu/easmith/
February 1, 2003                               Space Shuttle Columbia
Ad Astra Per Aspera                     To The Stars Through Asperity



More information about the EMBOSS mailing list