Error: unable to open file or unable to determine types for file bed file

Question

[sebastianzeki0]

I'm trying to do a bedtools compare on a couple of bed files. One of the bed files keeps throwing this error:

Error: unable to open file or unable to determine types for file

My file is as follows:

chr1     810865      3198369
chr1     844270      845356
chr1     882432      10373009
chr1     1104962     1173985
chr1     2616309     2617058
chr1     3056425     3245459
chr1     4704545     5447621

I have checked that the start position is before the end position. I have added and removed column headings chrom, "start" and "end" (also tried 'chromStart' and 'chromEnd'. But no joy. I had originally created this bed file from a tab delimited file from R. Does anyone know what has gone wrong?

[Ashutosh Pandey]

Try: perl -p -i -e 's/ /\t/g' bed_file.txt for converting white spaces to tabs and rerun bedtools.

[Devon Ryan]

Have you ensured that you're always using tabs to separate the columns? I've seen a couple files with spaces randomly thrown in and that tends to break things.

[David W]

As others have suggested, this is most likely caused by the file having spaces in place of tabs. I note you mention this table was written in R, where the default separator for write.table is indeed a space. I have this function in my .Rprofile to make writing bioinformatic-sy table easier:

write.simple.table <- function(...){
    write.table(..., quote=FALSE, row.names=FALSE, col.names=FALSE, sep='\t')
}

[Neel]

Error: unable to open file or unable to determine types for file

Hi i am trying to run this commands bedtools intersect -a Ec_k12.gff -b target_genes.txt -f 0.5 -wa -wb > genes_with_coordinates.bed and i am getting this error even after following given steps regarding this error on biostars/google. please help me if anyone know how to solve it? and the error is -

Error: unable to open file or unable to determine types for file / target_genes.tx

- Please ensure that your file is TAB delimited (e.g., cat -t FILE).
- Also ensure that your file has integer chromosome coordinates in the
  expected columns (e.g., cols 2 and 3 for BED).

thank you !

Answer 1

[sebastianzeki0]

OK so this is what I did in case anyone else gets the same problem

1. As per Ashutosh I ran perl -p -i -e 's/ /\t/g' I think this was not the problem though in the end but is probably a good idea
2. Uploaded the bed file to UCSC genome browser as a custom track. If the bed file is incorrectly formatted it will tell you what the problem is and is much better at debugging than bedtools
3. beware of directory paths when running the command line.

Sorted now

Thanks everyone

[Tim Amos]

I got the same error message from intersectBed using a gtf file.

I uploaded it to UCSC genome browser as a custom track as suggested (https://genome.ucsc.edu/cgi-bin/hgCustom) and it told me what my issue was:

chromStart after chromEnd (1462431 > 1462417)

Thanks

Note: I had to change my chromosome name to get this to work as E. coli isn't one of their species options.

[boczniak767]

Sometimes it is caused by header added by program which generated bed file. In this case tail -n +2 file.bed > newfile.bed helps

Answer 2

[Alex Reynolds]

Run cat -T on your file in order to verify if it has tabs where you think they should be. If it has multiple spaces where a tab is expected, use sed 's/ \+/\t/g' on the file to replace one or more spaces with a single tab character.

Answer 3

[swbarnes2]

One easy thing to try, despite the bed format only requiring those 3 columns, I've gotten that error, and fixed it by adding a fourth column with '+', because the command I was using need that information.