error using bedtools closestbed

Question

[chxu02]

a.bed looks like this:

chr1 631977 631979 -2.777777778
chr1 631994 631996 31.5625
......

b.bed looks like this:

chr1 11873 11874 DDX11L1 0 +
chr1 17435 17436 MIR6859-3 0 -
......

After I ran:

closestBed -a a.bed -b b.bed -D b > c.bed

c.bed looks like this:

chr1 631977 631979 -2.777777778 . -1 -1 . -1 . -1
chr1 631994 631996 31.5625 . -1 -1 . -1 . -1
......

Both a & b were formatted using galaxy before running. Help!

[chxu02]

Thanks guys. The spaces were inappropriately added by awk:

cat b.txt | awk '{if($3 == "-") {print $2, "\t", $5-1, "\t", $5, "\t", $3} else {print $2, "\t", $4, "\t", $4+1, "\t", $3} }' > b.bed

There were no spaces in b.txt. Any suggestion?

[TriS]

Can we have a snapshot of the b.txt file? You can edit your answer and add it there.

[marina.v.yurieva]

No commas in awk (your script puts spaces (comma) and tabs but you need only tabs:

cat b.txt | awk '{if($3 == "-") {print $2 "\t" $5-1 "\t" $5, "\t" $3} else {print $2 "\t" $4 "\t" $4+1 "\t" $3} }' > b.bed

Answer 1

[marina.v.yurieva]

It gave me this result when I ran your example:

chr1    631977  631979  -2.777777778    chr1    17435   17436   MIR6859-3       0       -       -614542
chr1    631994  631996  31.5625 chr1    17435   17436   MIR6859-3       0       -       -614559

Check that the files are tab-delimeted, there is something wrong with formatting. It should work.

Answer 2

[TriS]

Try 3 things:

1) Instead of:

closestBed -a a.bed -b b.bed -D b > c.bed

use

closestBed -a a.bed -b b.bed -D "b" > c.bed

2) Check that your file is formatted properly (are marina suggested):

head a.bed
head b.bed

if you don't get something like what you put above then do:

cat a.bed | tr '\r' '\n' > a_v2.bed #removes carriage returns and makes a new line

or if it's not tab delimited

cat a.bed | tr '--> x <--' '\t'  #where --> x is your separator is, i.e. ' ' (space) ',' (comma) etc...

3) Sort the file

sort -k1,1 -k2,2n a.bed > a_sorted.bed

--->> EDIT <<---

I just copy/pasted your input which looks like space delimited, I converted it to tab delimited:

cat a.bed | tr ' ' '\t' > a2.bed
cat b.bed | tr ' ' '\t' > b2.bed

then used closestBed:

closestBed -a a2.bed -b b2.bed -D "b"

which gave me

chr1    631977    631979    -2.777777778    chr1    17435    17436    MIR6859-3    0    -    -614542
chr1    631994    631996    31.5625    chr1    17435    17436    MIR6859-3    0    -    -614559

Answer 3

[chxu02]

Thanks. It's solved.

The b.txt looks like

NM_001276352 chr1 - 67092175 67134971 C1orf141
NM_000299 chr1 + 201283451 201332993 PKP1
......

When I ran

cat b.txt | awk '{if($3 == "-") {print $2 "\t" $5-1 "\t" $5 "\t" "\t" "\t" $3} else {print $2 "\t" $4 "\t" $4+1 "\t" "\t" "\t" $3} }' > b.bed

The result was properly printed. But when I ran

cat b.txt | awk '{if($3 == "-") {print $2 "\t" $5-1 "\t" $5 "\t" $6 "\t" $1 "\t" $3} else {print $2 "\t" $4 "\t" $4+1 "\t" $6 "\t" $1 "\t" $3} }' > b.bed

The last 2 columns went to second line for each original row

chr1 67134970 67134971 C1orf141
NM_001276352 -
chr1 201283451 201283452 PKP1
NM_000299 +