SEQanswers (
-   Bioinformatics (
-   -   read length a problem when mapping with MAQ (

koustavpal 08-27-2012 10:28 PM

read length a problem when mapping with MAQ

I have a file containing 36.8 million reads, as suggested in the MAQ manual I divided the file so that i had files containing 2 million reads. the minimum length of any read is 49bp, and the maximum i cannot approximate because whenever i input the files, if one file has reads with length greater than 65bp then maq returns a "no reasonable reads found" error.

I used the dynamic trim script to trim down the sequences according to their quality scores, which shortened the reads which before were 65bp long to 49-55bp long. These reads when inputted were promptly accepted and yet again when i input files containing sequences with length more than 65bp after trimming these files were rejected with the no reasonable reads found error.

This leads me to believe that maq may have an maximum read length cap of 60bp?

It would be helpful if someone can shed some more light on the matter.

swbarnes2 08-27-2012 10:32 PM

MAQ is really old. No one uses it anymore, it's not a fast algorithm. Switch to a Burrows-Wheeler Transform algorithm, like bwa or Bowtie.

koustavpal 08-27-2012 11:44 PM

I have used bowtie2 to achieve the same, but our lab head is persistent on doing the same with maq. Therefore, i must use maq.

HESmith 08-28-2012 06:05 AM

The MAQ FAQ states that the maximum read length is 63bp. Also, it will only align reads of the same length (although you can merge different read lengths post-alignment).

All times are GMT -8. The time now is 09:56 PM.

Powered by vBulletin® Version 3.8.9
Copyright ©2000 - 2020, vBulletin Solutions, Inc.