bedToBam gives empty output

[rivered]

Hi all,

I am trying to convert a .bed file which I made from data from an ancient genome tiling microarray into a .bam file. I am running into this annoying error which I do not seem able to resolve.

Running the following command:

bedtools bedtobam -i ~$/etc/file.bed -g ~$/etc/arabidopsis.txt > file.bam

I get a bam output file which is only 1 Kb. (For a point of reference my bed file is 108mb). The .bed file is a self-made BED5 format which looks like this:

Chr1 25 50 id-1 0.005
Chr1 60 85 id-2 0.001
Chr1 113 138 id-3 0.001
Chr1 154 179 id-4 0.359
Chr1 185 210 id-5 0.001
etc, etc etc


The genome file looks like this
Chr1 30427671
Chr2 19698289
Chr3 23459830
Chr4 18585056
Chr5 26975502
(Chrmt and chrCl are not necessary since these are not in the bed either)

When I head the 1kb file from the output this follows:
~/some_directory$ samtools view -H file.bam
@HD VN:1.0 SO:unsorted
@PG ID:BEDTools_bedToBam VN:Vv2.16.2
@SQ SN:Chr1 AS:/directories/arabidopsis.txt LN:30427671
@SQ SN:Chr2 AS:/directories/arabidopsis.txt LN:19698289
@SQ SN:Chr3 AS:/directories/arabidopsis.txt LN:23459830
@SQ SN:Chr4 AS:/directories/arabidopsis.txt LN:18585056
@SQ SN:Chr5 AS:/directories/arabidopsis.txt LN:26975502

It seems like the SQ and SN lines are generated and then it stops. Any ideas? All help will be greatly appreciated!

-R

[rivered]

Solved the problem!

Somehow R converted the generated .BED files into Macintosh(CR) dataformat. Checked eol conversion in notepad ++ and saw it was not UNIX format.

Then converted the files with the following command:
perl -p -e 's/\r/\n/g' < macfile.txt > unixfile.txt

Hope it helps someone,
R