errors in rules delly, manta and smoove

[warthmann]

Hello,
below command fails with errors in all relevant rules (delly, manta, smoove). A pretended run (-n) returns no problem, and I would appreciate help in getting it to run. I am attaching zipped archive containing the snakmake log file, config.yaml, sampleKey.txt, and a log of STDERR and STDOUT.
snakeSV-troubleshoot.zip

$snakeSV --configfile config/config.yaml --cores 8 -pr --use-conda

thanks a lot for your help
best
Norman

[ricardovialle]

Hi Norman,

Thanks for reaching out. I didn't test the pipeline with non-human genomes.
After a quick look here, manta also gave me errors, but I was able to make it run using dysgu instead.
The other tools: smoove and delly, use a file to restrict regions that are specific to human genome builds, so right now they will require some changes in the snakeSV rule to test without that option.

I recommend interpreting these results cautiously since I'm unsure if the results are reliable or even if these callers should work with non-human genomes.

Here is my config.yaml that worked:

# All paths set on this file are relative to the workdir (if not absolute).
workdir: "./"

OUT_FOLDER: "results_test"
SAMPLE_KEY: "sampleKey.txt"

TOOLS:
#  - "manta"
#  - "smoove"
#  - "delly"
  - "dysgu"

# Reference genome files
REFERENCE_FASTA: "Baxijiao.AAA.genome_plus_chloroplast.fa"
REF_BUILD: ""

Best,
Ricardo

[warthmann]

Hello @ricardovialle , thank you for the swift reply. Yes, my fear was that this is yet another tool that is human genomes only. However, I have run delly on this data before and generated the necessary files, so maybe it does not take much to make it more generic? Meanwhile, I did as instructed and reran specifying only the tool "dysgu". It seems to have gotten quite far, but did throw an error eventually. I am attaching the log. Any hint or fix would be greatly appreciated. Thanks a lot!
snakesv_2023-11-23_13-39-07.log

[ricardovialle]

Hi Norman,

I tested removing the dependencies and got to make it work with delly and dysgu

You can try running using this branch:
git clone -b testing_nonhuman_ref https://github.com/RajLabMSSM/snakeSV.git

You can run it under the same conda environment you already have. Just make sure to run the pipeline with the snakeSV executable cloned from the git repo.

Here is how my config.yaml looks like:

OUT_FOLDER: "results_test"
SAMPLE_KEY: "sampleKey.txt"

TOOLS:
  - "delly"
  - "dysgu"

# Reference genome files
REFERENCE_FASTA: "Baxijiao.AAA.genome_plus_chloroplast.fa"
REF_BUILD: ""