"Pathologically long" line error with legitimate structural variants

81 views
Skip to first unread message

user98

unread,
Apr 11, 2024, 3:10:13 PM4/11/24
to plink2-users
Hi, 

I'm working with a callset that contains SVs, very few of which are very long. I keep hitting the error "Pathologically long" line, even though these are legitimate variants. 

Here is a log:

PLINK v1.90b3.44 64-bit (17 Nov 2016)      https://www.cog-genomics.org/plink2
(C) 2005-2016 Shaun Purcell, Christopher Chang   GNU General Public License v3
Logging to callset_EUR_with_cms_chr2.oxford.inf.10kb.cM.log.
Options in effect:
  --cm-map genetic_map_hg38_withX_chr2.txt 2
  --keep-allele-order
  --make-bed
  --map callset_EUR_with_cms_chr.2.inf.10kb.bim
  --out callset_EUR_with_cms_chr2.oxford.inf.10kb.cM
  --ped callset_EUR_with_cms_chr.2.ped
  --zero-cms

257564 MB RAM detected; reserving 128782 MB for main workspace.
Error: Line 907969 of .map file is pathologically long.


Is there any way at all to tell plink to just ignore this and continue? 

Thanks, 

Christopher Chang

unread,
Apr 11, 2024, 3:32:25 PM4/11/24
to plink2-users
No, plink does not support variant IDs of length >16000; if you have a .map file with longer IDs, you are responsible for shortening them before providing them to plink.

Christopher Chang

unread,
Apr 11, 2024, 3:34:18 PM4/11/24
to plink2-users
(To be clear, the structural variants may be legitimate, but the variant IDs that have been assigned to them are not legitimate and must be fixed.)

user98

unread,
Apr 11, 2024, 3:58:24 PM4/11/24
to plink2-users
The variant IDs have already been shortened, it seems it's the alleles themselves in the .bim that plink is complaining about (I checked line 907969 manually). Strangely, I haven't run into this issue when generating the bed/bim/fam from a vcf or adding cM to the bims from a genetic map. 

Christopher Chang

unread,
Apr 11, 2024, 4:04:28 PM4/11/24
to plink2-users
Well, the .log you posted referred to a .map file so I tried to help with that.  If your real problem is with a .bim, I need to see a .log file which exhibits it.

plink 1.9 and 2.0 are supposed to be able to handle allele codes of length up to a few tens of millions.

Christopher Chang

unread,
Apr 11, 2024, 4:06:19 PM4/11/24
to plink2-users
...oh, I now see that you provided a .bim file to --map.

Try generating an actual .map file that contains only the first four columns of that .bim.  "cat [input .bim file] | cut -f 1-4 > [new .map filename]"

Javier Maravall

unread,
Apr 11, 2024, 4:27:18 PM4/11/24
to Christopher Chang, plink2-users
That worked, thanks a lot!

--
You received this message because you are subscribed to a topic in the Google Groups "plink2-users" group.
To unsubscribe from this topic, visit https://groups.google.com/d/topic/plink2-users/ezhCka-veuY/unsubscribe.
To unsubscribe from this group and all its topics, send an email to plink2-users...@googlegroups.com.
To view this discussion on the web visit https://groups.google.com/d/msgid/plink2-users/0cc4e1f8-8b37-43f3-b13e-863e5ef28fc4n%40googlegroups.com.
Reply all
Reply to author
Forward
0 new messages