Bedops --difference errors

Question

[rbronste]

Hi

bedops --difference giving me some errors in the final bed file and Im wondering why, how to get a workaround?

chr8 9 69893500 5.00E+07 95000416 65834

Getting some intervals that look like this and not sure where its going off the rails when trying to subtract a genomic space from another interval file? Thanks!

Rob

[Alex Reynolds]

Are your inputs sorted (i.e., with sort-bed)? Do you have files with Windows line endings (e.g., from Excel spreadsheets)? If you want to post your files somewhere, I'm happy to take a quick look.

[rbronste]

I tried to sort immediately following bedops --difference with a file of CTCF coordinates in mm10 however get the following error message right after when I try to sort:

Error on line 67243 consensus_bedops_CTCFdiff.bed. Genomic end coordinate is less than (or equal to) start coordinate.

[Alex Reynolds]

You should sort your input files (regions of interest and CTCF regions) with sort-bed before you do any set operations with bedops. You only need to sort once, but you need to make sure the inputs are sorted before you work on them.

[rbronste]

Bedops sorting has been weird for me in general, often get this sort of error message:

BED row length exceeds capacity at line 1 in consensusPeaks_DiffBind_mm10.bed. Check that you have unix newlines (cat -A) or increase TOKENS_MAX_LENGTH in BEDOPS.Constants.hpp and recompile BEDOPS.

[Alex Reynolds]

Are you working with Windows or Microsoft Office-sourced files? This can cause problems with Unix-based bioinformatics tools of all kinds, including BEDOPS. Here's one way to strip Windows newlines and sort the BED file in one pass:

$ tr -d '\r' < bad.bed | sort-bed - > good.bed

[rbronste]

Very infrequently I open something in excel but primarily work in through command line on a computing cluster. Though yes I may have stupidly polluted some of these. Thanks for the tip!