ready for use?

[nextgenusfs]

is this tool ready for use? We'd really like to give this model a try as we have lots of data where we have both Illumina and ONT data. It won't run from an existing functional clair3 environment, multiple command line options are not getting parsed properly due to the multiple layers of shell scripts. For example,

(clair3) [jmp9484@biolnxlogin01-prd mp-trial]$ cat mp_out/run_clair3_mp.log 
[INFO] BAM 1 FILE PATH: ./ont-aligns.bam
[INFO] BAM 2 FILE PATH: ./illumina-aligns.bam
[INFO] BAM 1 PLATFORM: ont
[INFO] BAM 2 PLATFORM: ilmn
[INFO] REFERENCE PATH: /path/to/ref.fasta
[INFO] THREADS: 1
[INFO] CLAIR3 PILE UP MODEL PATH FOR BAM 1: /path/to/models/clair3_models/r1041_e82_400bps_hac_v420
[INFO] CLAIR3 PILE UP MODEL PATH FOR BAM 2: /path/to/models/clair3_models/ilmn
[DEPRECATED]  CLAIR3 PILE UP PREFIX: pileup
[INFO] CLAIR3-TRIO MODEL PATH: /path/to/models/clair3_mp_models/ont_ilmn
[INFO] CLAIR3-TRIO MODEL PREFIX: variables
[INFO] OUTPUT FOLDER: ./mp_out
[INFO] BED FILE PATH: EMPTY
[INFO] VCF FILE PATH: EMPTY
[INFO] CONTIGS: EMPTY
[INFO] BAM 1 NAME:   sample1_ont
[INFO] BAM 2 NAME: sample1_ilmn
[INFO] CHUNK SIZE: 5000000
[INFO] SAMTOOLS PATH: /miniconda/path/envs/clair3/bin/samtools
[INFO] PYTHON PATH: /miniconda/path/envs/clair3/bin/python3
[INFO] PYPY PATH: pypy3
[INFO] PARALLEL PATH: /miniconda/path/envs/clair3/bin/parallel
[INFO] WHATSHAP PATH: /miniconda/path/envs/clair3/bin/whatshap
[INFO] QUALITY THRESHOLD: 2
[INFO] FULL ALIGN PROPORTION: 0.3
[INFO] FULL ALIGN REFERENCE PROPORTION FOR BAM 1: 0.1
[INFO] FULL ALIGN REFERENCE PROPORTION FOR BAM 2: 0.3
[INFO] PHASING PROPORTION FOR BAM 1: 0.7
[INFO] PHASING PROPORTION FOR BAM 2: 0.7
[INFO] ENABLE FILEUP ONLY CALLING: False
[INFO] ENABLE PILEUP CALLING AND PHASING: False
[INFO] ENABLE FAST MODE CALLING: False
[INFO] ENABLE CALLING SNP CANDIDATES ONLY: False
[INFO] ENABLE PRINTING REFERENCE CALLS: False
[INFO] ENABLE OUTPUT GVCF: False
[INFO] ENABLE HAPLOID PRECISE MODE: False
[INFO] ENABLE HAPLOID SENSITIVE MODE: False
[INFO] ENABLE INCLUDE ALL CTGS CALLING: True
[INFO] ENABLE NO PHASING FOR FULL ALIGNMENT: True
[INFO] ENABLE REMOVING INTERMEDIATE FILES: False
[INFO] ENABLE PHASING VCF OUTPUT: False
[INFO] ENABLE LONG INDEL CALLING: False
./mp_out
[INFO] You are running ont + ilmn hybrid.
[INFO] * Clir3-Trio pipeline start
[INFO] * 0/7 Check environment variables
[INFO] --include_all_ctgs enabled
[WARNING] Please enable --no_phasing_for_fa if calling variant in non-diploid organisms
[ERROR] whatshap not found, please check you are in clair3 virtual environment
[ERROR] Current python execution path: /miniconda/path/envs/clair3/bin/python

Since most of the codebase seems to be in python, honestly I don't quite understand the multiple layers/levels of shell scripts...

[sujunhao]

Hello,

Thank you for your message. The tool is ready for running.

Based on your message, it appears that there may be an installation issue during your testing, as the log indicated that the Clair3-MP was unable to find whatshap in your install path.

As the Clair3-MP environment is slightly different from Clair3, installing all packages in a new environment is recommended when running Clair3-MP. To resolve the issue in your log file, could you please try to reinstall Clair3-MP using the instructions provided on the README page (https://github.com/HKU-BAL/Clair3-MP#option-2-anaconda-install)?

On the other hand, we added docker support in version v1.0, which you may find more convenient at testing.

I hope this helps. Let me know if you have any further questions or concerns.

Regards,
Junhao

[nextgenusfs]

Its hard to imagine its an environment issue, whatshap is clearly in the PATH of the environment and I get the same error if I pass --whatshap=$(which whatshap), but even more frustrating is that we are working with haploid organism and I think whatshap isn't even going to be used.

So since Clair3-MP seems to require regular Clair3 models are you saying that the versions are not compatible? One cannot use the tools in the same environment to call just ONT data with Clair3 and ONT-Illumina with Clair3-MP?

[sujunhao]

Hi,

Yes, Clair3 and Clair3-MP environment is not compatible. Clair3-MP uses an old version of Clair3 and uses python3.6 for running, while Clair3's newest version is updated to support python3.9. To run Clair3-MP, please try to reinstall a Clair3-MP environment.

Clair3-MP added Singularity support now. It may be more convenient to use singularity for testing.

I hope this can help.

Thanks,
Junhao