Buffer overflow detected

[jbreynier]

Hello,
Thank you very much for providing this tool. I'm running TRUST4 on 3' scTCR-seq samples, here is the command I used (for sample sn09):
run-trust4 -u /data/sn09*_R2_001.fastq.gz --barcode /data/sn09*_R1_001.fastq.gz --UMI /data/sn09*_R1_001.fastq.gz --readFormat bc:0:15,um:16:-1 --repseq -f /ref/human_IMGT+C.fa --ref /ref/human_IMGT+C.fa -t 4 -o /output/sn09

It works very well for all my samples, except for one sample which throws a buffer overflow error at the annotation step:
[Mon Nov 20 17:06:26 2023] SYSTEM CALL: /usr/local/bin/annotator -f /ref/human_IMGT+C.fa -a /output/sn09_final.out -t 4 -o /output/sn09 --barcode --UMI -r /output/sn09_assembled_reads.fa --airrAlignment > /output/sn09_annot.fa
[Mon Nov 20 17:06:26 2023] Start to annotate assemblies.
*** buffer overflow detected ***: /usr/local/bin/annotator terminated
sh: line 1: 2945762 Aborted (core dumped) /usr/local/bin/annotator -f /ref/human_IMGT+C.fa -a /output/sn09_final.out -t 4 -o /output/sn09 --barcode --UMI -r /output/sn09_assembled_reads.fa --airrAlignment > /output/sn09_annot.fa
system /usr/local/bin/annotator -f /ref/human_IMGT+C.fa -a /output/sn09_final.out -t 4 -o /output/sn09 --barcode --UMI -r /output/sn09_assembled_reads.fa --airrAlignment > /output/sn09_annot.fa failed: 34304 at /usr/local/bin/run-trust4 line 54.

Both sn09_assembled_reads.fa (3.6G) and sn09_final.out (473M) look fine to me. I have the same issue using Singularity on my local cluster or Docker on an AWS instance. Have you encountered this before/have any suggestions on how to fix? I'm currently using the latest Docker image (1.0.13--h43eeafb_0). Thank you in advance!

[mourisl]

This could be a bug in TRUST4. Could you please share the sn09_final.out with me? Thank you.

[jbreynier]

Sent by email, thank you for your help.

[mourisl]

Thank you for sharing the file. I think I have fixed the bug and uploaded the new code to github. Could you please clone the github version and give it a try?

[jbreynier]

The fix worked, thank you so much for the quick help!