"Error: piscem mapping failed with exit status ExitStatus(unix_wait_status(139))"

[wmacnair]

Hi simpleaf team

First off, a thank you: we've been using simpleaf to do lots of mapping for a single cell atlas project, and it has been making everything faster and smoother 🚀 (especially being able to S/U/A counts super fast). This is awesome 😄

However, there are a small proportion of samples where it doesn't work, and we can't really see any reason why. I've pasted some example output below.

This doesn't appear to driven by e.g. tiny file sizes due to RNAseq not working:

-rw-r----- 1 macnairw games 1.2G Jul 23 21:20 WM177_POOL4_S9_L001_I1_001.fastq.gz
-rw-r----- 1 macnairw games 3.8G Jul 23 21:21 WM177_POOL4_S9_L001_R1_001.fastq.gz
-rw-r----- 1 macnairw games 8.5G Jul 23 21:22 WM177_POOL4_S9_L001_R2_001.fastq.gz
-rw-r----- 1 macnairw games 1.2G Jul 23 21:22 WM177_POOL4_S9_L002_I1_001.fastq.gz
-rw-r----- 1 macnairw games 3.7G Jul 23 21:22 WM177_POOL4_S9_L002_R1_001.fastq.gz
-rw-r----- 1 macnairw games 8.4G Jul 23 21:23 WM177_POOL4_S9_L002_R2_001.fastq.gz

Right now it's difficult to diagnose what is going on... Most samples in this experiment work, a handful don't, and I don't really see anything I can work with from the error message. Any suggestions??

Thanks! Will

(Tagging @marusakod as she has had similar problems.)

simpleaf --version
# simpleaf 0.14.1

# simpleaf configuration
export ALEVIN_FRY_HOME="/projects/site/pred/neurogenomics/resources/scprocess_data/data/alevin_fry_home"
simpleaf set-paths

# mess about with input strings
R1_fs=$(echo ./Macnair_2022/data/fastqs/WM177_POOL4_S9_L001_R1_001.fastq.gz ./Macnair_2022/data/fastqs/WM177_POOL4_S9_L002_R1_001.fastq.gz | sed "s/ /,/g")
R2_fs=$(echo ./Macnair_2022/data/fastqs/WM177_POOL4_S9_L001_R2_001.fastq.gz ./Macnair_2022/data/fastqs/WM177_POOL4_S9_L002_R2_001.fastq.gz | sed "s/ /,/g")

# simpleaf quantfication
simpleaf quant \
  --reads1 $R1_fs \
  --reads2 $R2_fs \
  --threads 16 \
  --index $ALEVIN_FRY_HOME/human_2020-A_splici/index \
  --chemistry 10xv3 --resolution cr-like \
  --expected-ori fw \
  --t2g-map $ALEVIN_FRY_HOME/human_2020-A_splici/index/t2g_3col.tsv \
  --unfiltered-pl $CELLRANGER_DIR/3M-february-2018.txt --min-reads 1 \
  --output ./Macnair_2022/output/ms01_alevin_fry/af_WM177

# 2023-07-24T08:42:52.328031Z  INFO simpleaf::simpleaf_commands::quant: piscem map-sc cmd : /projects/site/pred/neurogenomics/resources/scprocess_data/conda/af/bin/piscem map-sc --index /projects/site/pred/neurogenomics/resources/scprocess_data/data/alevin_fry_home/human_2020-A_splici/index/piscem_idx --threads 8 -o ./Macnair_2022/output/ms01_alevin_fry/af_WM177/af_map -1 ./Macnair_2022/data/fastqs/WM177_POOL4_S9_L001_R1_001.fastq.gz,./Macnair_2022/data/fastqs/WM177_POOL4_S9_L002_R1_001.fastq.gz -2 ./Macnair_2022/data/fastqs/WM177_POOL4_S9_L001_R2_001.fastq.gz,./Macnair_2022/data/fastqs/WM177_POOL4_S9_L002_R2_001.fastq.gz --geometry chromium_v3
# 2023-07-24T08:43:31.448195Z ERROR simpleaf::utils::prog_utils: command unsuccessful (signal: 11 (SIGSEGV) (core dumped)): "/projects/site/pred/neurogenomics/resources/scprocess_data/conda/af/bin/piscem" "map-sc" "--index" "/projects/site/pred/neurogenomics/resources/scprocess_data/data/alevin_fry_home/human_2020-A_splici/index/piscem_idx" "--threads" "8" "-o" "./Macnair_2022/output/ms01_alevin_fry/af_WM177/af_map" "-1" "./Macnair_2022/data/fastqs/WM177_POOL4_S9_L001_R1_001.fastq.gz,./Macnair_2022/data/fastqs/WM177_POOL4_S9_L002_R1_001.fastq.gz" "-2" "./Macnair_2022/data/fastqs/WM177_POOL4_S9_L001_R2_001.fastq.gz,./Macnair_2022/data/fastqs/WM177_POOL4_S9_L002_R2_001.fastq.gz" "--geometry" "chromium_v3"
# 2023-07-24T08:43:31.449914Z ERROR simpleaf::utils::prog_utils: stderr :
# ====
# 2023-07-24T08:42:52.373983Z  INFO piscem: cmd: ["sc_ref_mapper", "-i", "/projects/site/pred/neurogenomics/resources/scprocess_data/data/alevin_fry_home/human_2020-A_splici/index/piscem_idx", "-g", "chromium_v3", "-1", "./Macnair_2022/data/fastqs/WM177_POOL4_S9_L001_R1_001.fastq.gz,./Macnair_2022/data/fastqs/WM177_POOL4_S9_L002_R1_001.fastq.gz", "-2", "./Macnair_2022/data/fastqs/WM177_POOL4_S9_L001_R2_001.fastq.gz,./Macnair_2022/data/fastqs/WM177_POOL4_S9_L002_R2_001.fastq.gz", "-t", "8", "-o", "./Macnair_2022/output/ms01_alevin_fry/af_WM177/af_map"]
# [2023-07-24 10:42:52.376] [info] loading index from /projects/site/pred/neurogenomics/resources/scprocess_data/data/alevin_fry_home/human_2020-A_splici/index/piscem_idx
# [2023-07-24 10:42:53.777] [info] done loading index
# processed (13000000) reads; (11168505) had mappings.====
# Error: piscem mapping failed with exit status ExitStatus(unix_wait_status(139))

[rob-p]

Hi @wmacnair and @marusakod,

The bioconda process how now completed as well and you should be able to pull piscem v0.6.1 from bioconda. Please let me know at your earliest convenience if this addresses all of the problem cases on your end.

Also, thank you again for your continued interaction and help in reproducing and tracking down this bug. It's users like yourselves that make it possible for us to work at building high-quality and robust software, and without your patience and support, that wouldn't be possible.

Edit: I checked with the bioconda version of piscem v0.6.1 with the shared data and index, but will wait for your verification just in case.

Best, Rob

[marusakod]

Hi @rob-p,

I pulled the new piscem version from bioconda and ran simpleaf with a bunch of different files where I previously got the error (including the ones that @wmacnair last shared with you) and I am happy to report that the error is gone!:)

Thank you so much for your help! if we encounter any errors in the future, we'll be happy to reach out.

Best, Maruša

[rob-p]

Hi @marusakod,

Thank you so much for confirming. That's great news. Thank you also for the detailed bug reports and data sharing. It really helped to track down (and ultimately fix) a bug that would have been so much trickier otherwise. Of course, please do reach out in the future if you encounter anything else unexpected or have any questions.

Thanks, Rob

[wmacnair]

Hi @rob-p

I've done a clean install of simpleaf and can also report that so far, files which previously failed have run fine 🥳

We have both very much appreciated your help here - simpleaf / the alevin ecosystem is fast and light, and we would very much like it to succeed!

Thanks again for your help Will