HKU-BAL / Clair3

Clair3 - Symphonizing pileup and full-alignment for high-performance long-read variant calling
246 stars 27 forks source link

Run does not respect AF settings #314

Closed Krannich479 closed 4 months ago

Krannich479 commented 4 months ago

Dear Clair3 dev team,

Background

I am using Clair3 for SNV/indel detection from ONT sequence data. The sequence data is from an amplicon sequencing experiment (high read coverage) of a mixed SARS-CoV-2 sample.

Clair3 call

run_clair3.sh \
    --bam_fn="${bam}" \
    --bed_fn="${params.bed}" \
    --ref_fn="${params.ref}" \
    --threads="${task.cpus}" \
    --platform="ont" \
    --model_path="${params.model}" \
    --snp_min_af=0.08 \
    --indel_min_af=0.15 \
    --no_phasing_for_fa \
    --var_pct_full=1 \
    --ref_pct_full=1 \
    --output="${baseDir}/${params.outdir}"

Issue

The run of Clair3 does not respect the set snp_min_af and indel_min_af. Does any combination of parameters here interfere or overwrite each other? In the merge_output.vcf.gz as well as the full_alignment.vcf.gz I see a line

NC_045512.2     22386   .       TTC     T       16.05   PASS    F       GT:GQ:DP:AD:AF  0/1:16:846:634,9:0.0106

which should not be present according to the parameters because AF=0.01.

Expected behavior

I expected to not have any variant records (at least not with a FILTER=PASS) in the VCF files that are of AF lower than 0.08 and 0.15 for SNPs and indels, respectively.

Additional info

Clair3 version: 1.0.8 Clair3 model: r1041_e82_400bps_hac_g632 OS: Ubuntu 20.04.6 LTS

aquaskyline commented 4 months ago

Please check if this addresses your concern #253.

Krannich479 commented 4 months ago

That does help explaining my findings. Thanks a lot!