intersection, Bed file and VCF.

Question

[jaqx008]

Hello All, I am trying to perform intersection between with my bed file and vcf and I keep getting errors.my bed files is a call of all the genes in the ORFs. The following is the command I am using and the errors.

bedtools intersect -a a.vcf -b b.bed -wa

error

• 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).

I tried to do cat -t a.vcf to TAB it but still get the errors. I will appreciate your suggestions. Thanks in advance.

Answer 1

[Alex Reynolds]

You might try BEDOPS bedops --ec:

$ bedops --ec --element-of 1 <(vcf2bed < a.vcf) <(sort-bed b.bed) > answer.bed

The use of --ec with bedops can return useful debugging information about malformed BED data, at the cost of a slower runtime. The --ec operator can be removed once the input is debugged.

[jaqx008]

Ok let me try this. Thanks

[jaqx008]

bedops --ec < a.vcf < b.bed > answer.bed

Error -bash: bedops: command not found

[Alex Reynolds]

You'll need to install it, probably: http://bedops.readthedocs.io/en/latest/content/installation.html

Answer 2

[dariober]

It seems the problem is not related to bedtools but rather with your input files.

cat -t a.vcf doesn't fix the problem, rather it tells you whether your columns are tab delimited. With cat -t a.vcf (or cat -t b.bed) you should see on screen columns separated by ^I. Perhaps your columns are separated by spaces instead? If so, you could fix it with tr ' ' '\t' < a.vcf > A.vcf

Also, as the error message suggests, have a look if you coordinates are actually integers. Try to do sort -k2,2n a.vcf to see if the top rows (excluding the header) have integers in column 2.

[jaqx008]

realized the error was in my bed file. I made sure all tabs are corrected and ran it again. but got complain about the one scaffold.

bedtools intersect -a a.vcf -b b.bed > file.txt Error: Invalid record in file genoOrfs.bed. Record is scaffold1090size18908_225 18674 18522

It seem fine like the rest to me. Don't know why the error

[dariober]

...Just in case you left out some space-to-TAB conversion, try to do

grep ' ' genoOrfs.bed # This shouldn't print any line.
grep 'scaffold1090size18908_225' genoOrfs.bed | cat -vet # Print the faulty record showing TABs

[jaqx008]

I tried it, the first command gave an output with Zero byte second gave

scaffold1090size18908_225^I18674^I18522$

Not sure what they mean..