UCSC-Tools hgWiggle out-of-memory error

32 views
Skip to first unread message

Ponomarev, Sergei (NIH/NIAID) [C]

unread,
Dec 6, 2016, 3:04:52 PM12/6/16
to gen...@soe.ucsc.edu, Gentzel, Amy (NIH/NIAID) [C], Klima, Robert (NIH/NIAID) [C], Yun, Alvin (NIH/NIAID) [C], Ponomarev, Sergei (NIH/NIAID) [C]
Dear UCSC Genome Browser Developer Team:

We are having issues with one of the UCSC-Tools/3.3.5 executables, the hgWiggle:

$ module load UCSC-Tools/3.3.5-goolf-1.7.20
$ cd Work/UCSC-Tools/additional_annotations/
$ hgWiggle -rawDataOut qual > qual.raw.wig
errAbort re-entered due to out-of-memory condition. Exiting.

$ pwd
/home/username/Work/UCSC-Tools/additional_annotations
[username@host additional_annotations]$ ls -ltrh
total 5.2G
lrwxrwxrwx 1 username username   11 Oct 24 15:51 qual.wig -> quality.txt
-rw-r----- 1 username username  12M Oct 24 15:51 simpleRepeat.txt.gz
-rw-r----- 1 username username  12M Oct 24 15:51 nestedRepeats.txt.gz
-rw-r----- 1 username username 2.4G Oct 24 15:51 qual.wib
-rw-r----- 1 username username 253M Oct 24 15:51 quality.txt
-rwxr-xr-x 1 username username  16M Dec  5 14:55 hgWiggle
-rw-r--r-- 1 username username    0 Dec  5 14:57 qual.raw.wig

I have been testing the latest pre-compiled 64-bit Linux versions of hgWiggle from


Thank you in advance!

Sincerely Yours,

Sergei Y. Ponomarev (NIH/NIAID) [C], PhD,  Contractor
Senior Systems Administrator - Linux and High Performance Computing Team
SRA International, Inc., A CSRA Company
Operations and Engineering Branch
Work phone: 301-761-6731, Mobile: 301-502-1673
5601 Fishers Lane, 6A66, Rockville, MD 20852
NIAID, National Institutes of Health, DHHS

Hiram Clawson

unread,
Dec 7, 2016, 1:50:42 PM12/7/16
to Ponomarev, Sergei (NIH/NIAID) [C], gen...@soe.ucsc.edu, Gentzel, Amy (NIH/NIAID) [C], Klima, Robert (NIH/NIAID) [C], Yun, Alvin (NIH/NIAID) [C]
Good Morning Sergei:

I'm curious about what you have for input files to this command.

Is the quality.txt file the file produced by the hgLoadWiggle command ?

Do you have the original ascii wiggle data before hgLoadWiggle constructed
the .wib file ? Depending upon the range of values to be encoded, the
values obtained from the .wib file can be different from the original
input values to hgLoadWiggle.

How many chromosomes or scaffolds are in this data ? Are any of them
unusually large ?

I can guess that hgWiggle could run out of memory depending upon your
shell user process memory limits if it is trying to get all data into
memory before it starts the output.

Does the command function if you specify one chromosome or scaffold
with the -chrom=<name> option ?

--Hiram
Reply all
Reply to author
Forward
0 new messages