Badly formed .bed file

[gormleymp]

I am trying to get mapping statistics for the number of reads matching to the target region in an exome capture experiment. I have a bed file from the library manufacturer. I have tried both Bedtools and BedUtils. Both files tell me the .bed file is not properly formatted asking if the start and end position are integers. The file appears to be properly formatted to me. Lines are tab-delimitted and look like:

chr1 728281 728500

Any thoughts?

Thanks,
Mike

[GenoMax]

Have you tried to upload the bed file to UCSC browser to see if it complains about the format?

Another option is to try suggestion in post #3 here: http://seqanswers.com/forums/showthread.php?t=17721

Did you open/edit the file on PC/Mac and then are trying to use it on unix?

[gormleymp]

UCSC does not complain. This is a file I downloaded directly from the vendor and use in UNIX only.

[aggp11]

Hi,

Could you list the exact "error message" that both BedTools and BedUtils are reporting? May be a line or two in your file has more columns than the others. It'll be easier to troubleshoot with the error message.

[GenoMax]

Are the chromosome identifiers the same in your reference file as compared to the bed file?

[gormleymp]

This error seems to have been caused by a windows to unix conversion as GenoMax suggested. Cat -v .bedfile revealed the hidden Windows ^M newline. dos2unix removed these characters and bamUtil is working now.

[aggp11]

GenoMax,

I think if the chromosome names don't match, bedtools neither returns an output nor reports an error. I am not sure about bamUtil as I have never used it before.