Malloc() when activating multimapping

[pietermonsieurs]

Hi Alex,
I have a problem which is very similar to the one described here: #1277 posted by @cnk113.
However, in contrast with the original post, my problem is not solved when running this with the latest version of STAR (2.7.10a_alpha_220314).

I have 3 different 10X datasets (3' gene expression), with 417M, 538M and 619M reads respectively. When running the code without multimapping, everything goes fine. When activating the multimapping (--soloMultiMappers Uniform), I only get an output for the first sample (417M reads). For the other two, I get an error message like:

STAR version: 2.7.10a_alpha_220314 compiled: 2022-03-30T12:27:55+0200 ln1.leibniz:/data/antwerpen/205/vsc20587/software/STAR/source
Apr 29 11:45:32 ..... started STAR run
Apr 29 11:45:39 ..... loading genome
Apr 29 11:45:46 ..... started mapping
Apr 29 12:35:49 ..... finished mapping
Apr 29 12:35:50 ..... started Solo counting
malloc(): invalid size (unsorted)

However, when lowering the number of reads via the --readMapNumber option to 400M, I got an output for the sample with 417M and 619M reads, but nog for the 538M reads sample (same error message as above). When further lowering to 200M reads, I got a valid output for all three samples.

Any idea how to solve this issue?

The command I use is:
$star_bin
--readFilesCommand zcat
--readFilesIn $fastq_file_2 $fastq_file_1
--runThreadN $threads
--soloType CB_UMI_Simple
--soloCBwhitelist $bc_whitelist
--soloCBstart 1 --soloCBlen 16
--soloUMIstart 17 --soloUMIlen 12
--soloFeatures Gene
--genomeDir $genome_index_dir
--soloMultiMappers Uniform
--outFileNamePrefix $out_prefix
--alignIntronMax 1
--outSAMtype BAM Unsorted
--soloCellFilter EmptyDrops_CR 16000 0.99 10 45000 90000 500 0.01 20000 0.01 10000

Best,
Pieter

[alexdobin]

Hi Pieter,
could you send me Log.out file for one of the failed runs?
Thanks!
Alex

[pietermonsieurs]

Hi Alex,
I have uploaded the Log.out for the three samples I run. Only sample01 runs without error, sample02 and sample03 crash when starting the counting (before collapsing the UMIs).
Thanks!
Pieter

Log.sample03.out.txt
Log.sample01.out.txt
Log.sample02.out.txt

[alexdobin]

Hi Pieter,

thanks for the files! Nothing suspicious in the Log files.
I tried to reproduce the error using your parameters on my test datasets, but did not get any errors.
Please try this patch:
https://github.com/alexdobin/STAR/releases/tag/2.7.10a_alpha_220506
It fixes some issues with multimappers, but I am not sure if it will help in your case.

Thanks!
Alex

[pietermonsieurs]

Hi Alex,
Thanks for the response!
Unfortunately, I still get the same response with the newest patch (one sample works, the two other fail when starting the counting).
Best regards,
Pieter

[alexdobin]

Hi Pieter,

have you seen this problem on any public dataset?
If not, I am afraid to reproduce and fix the issue I will need the full dataset on which STARsolo fails.

[pietermonsieurs]

Hi Alex,
Sure, I can share the data. Can I send you a download link via mail?
Many thanks!
Pieter

[alexdobin]

Hi Pieter,

sure, my email is dobin @ cshl edu

Thanks!
Alex

[alexdobin]

Hi Pieter,

thanks for the files. I could reproduce and the problem and am working to fix it.

Cheers
Alex

[pietermonsieurs]

Hi Alex,
Great! Thanks for all the efforts.
Let me know when you need more information.
Cheers,
Pieter

[alexdobin]

Hi Pieter,

I think I have fixed this bug, please try this patch:
https://github.com/alexdobin/STAR/releases/tag/2.7.10a_alpha_220601

Thanks!
Alex

[pietermonsieurs]

Hi Alex,
Awesome, it works!! (for both the samples that failed in my explanation above, I now get a correct out (raw and filtered matrix).
Thank you very much for all the efforts!
Cheers,
Pieter

[alexdobin]

Hi Pieter,

Great! Many thanks for helping to debug this issue!

Cheers
Alex

[alexlenail]

I also had this issue, and the patch also fixed my issue.