You signed in with another tab or window. Reload to refresh your session.You signed out in another tab or window. Reload to refresh your session.You switched accounts on another tab or window. Reload to refresh your session.Dismiss alert
I found a broken BED record is printed by intersectBed if it is very long (probably longer than 1024bytes).
Here is an example:
# envirionment
% uname -sro ; bedtools --version
Linux 3.10.0-1160.el7.x86_64 GNU/Linux
bedtools v2.30.0
# ttn.bed12 has a single BED12 formatted transcript record
% cat ttn.bed12 | wc
1 12 3319
# the BED12 formatted record become broken as 11 columns
% intersectBed -wa -a ttn.bed12 -b ttn.bed3 | wc
1 11 1024
I expected to have 12 column BED record output in the last command, but the output was broken. It seems to occur when a record is longer than 1024 bytes. Notably, 'bed12ToBed6' handle the record properly and I guess it might
be an issue in printing of intersectBed, rather than the fundamental difficulties of the BEDtools suite.
The full data content of this example is below. Note that this is an actual record of human gene, not artificial example.
A naive workaround for the issue reported in arq5x#1049.
The size of printing buffer is increased from 1024 to 8192 to handle very long entry, for example a gene with exceptionally many exons.
Hi,
I found a broken BED record is printed by
intersectBed
if it is very long (probably longer than 1024bytes).Here is an example:
I expected to have 12 column BED record output in the last command, but the output was broken. It seems to occur when a record is longer than 1024 bytes. Notably, 'bed12ToBed6' handle the record properly and I guess it might
be an issue in printing of
intersectBed
, rather than the fundamental difficulties of the BEDtools suite.The full data content of this example is below. Note that this is an actual record of human gene, not artificial example.
The text was updated successfully, but these errors were encountered: