search

Illumina / Pisces

Somatic and germline variant caller for amplicon data. Recommended caller for tumor-only workflows.
GNU General Public License v3.0
92 stars 16 forks source link

Pisces has different dp and no call？ #68

Open user-tq opened 2 years ago

user-tq commented 2 years ago

I'm learning the somatic mutation process, Out of curiosity, I compared the results of muect2 and Pisces, and I found that there would be a great depth difference between them. my cmd

/Data2/zhangjs/data/soft/fastp   -w 8 -5 -3 -q 20 -n 3 -l 36 -i $raw_fq1  -o $clean_R1.fq.gz   -I $raw_fq2  -O   $clean_R2.fq.gz
/home/tanqiang/bwa_mem2/bwa-mem2-2.2.1_x64-linux/bwa-mem2 mem  -t 8 -M -R  "@RG\tID:$sample_name\tPL:Illumina\tLB:$sample_name\tSM:$sample_name\tPG:bwa"  $hg19.fasta    $clean_R1.fq.gz   $clean_R2.fq.gz |samtools view -b -o  $name.bam
sambamba sort -m 8G -t 8  $name.bam  -o  $name.sort.bam
sambamba markdup -t 8  $name.sort.bam   $name.dedup.sort.bam

Pisces -b   $name.dedup.sort.bam   -g pisces_ref/  -i $target_bed --outfolder  pisces_vcf/   -VariantQualityFilter 20  --minbq 20 --minmq 30 --minvf 0.001   --filterduplicates true

gatk  Mutect2  --minimum-mapping-quality 30   --min-base-quality-score 20  --minimum-allele-fraction   0.001 -R  $hg19.fasta -I  $name.dedup.sort.bam   -tumor $name   -L $target_bed -O $name_no_filter_tag.mutect2.vcf

in pisces_vcf 

chr7    140494073   .   G   .   100 PASS    DP=1901 GT:GQ:AD:DP:VF:NL:SB    0/.:100:1895:1901:0.0032:20:-100.0000
chr7    140494074   .   T   .   100 PASS    DP=1956 GT:GQ:AD:DP:VF:NL:SB    0/.:100:1924:1956:0.0164:20:-100.0000

in mutect2_vcf

chr7    140494073   .   G   GT  .   .   AS_SB_TABLE=917,593|12,9;DP=1590;ECNT=1;MBQ=20,35;MFRL=194,204;MMQ=60,60;MPOS=16;POPAF=7.30;RPA=9,10;RU=T;STR;TLOD=9.65 GT:AD:AF:DP:F1R2:F2R1:SB    0/1:1510,21:0.015:1531:568,9:454,8:917,593,12,9

I'm confused: (1).Why are they different in depth?(1901,1531) (2).I'm trying to use same options,but pisces doesn't seem to think 'G>GT' is right ? i make a mini bam

samtools view bam/jm_lghong.dedup.sort.bam  "chr7:140494000-140494100

mini.sam.gz

Thank you in advance

tamsen commented 2 years ago

Hi there,

Thanks for your interest.

Pisces has moved here, so please make sure you are using the latest version, and post questions here https://github.com/tamsen/Pisces in the future.

Pisces and mutect have different filters on which basecalls they will accept. By default Pisces does not use basecalls with basecall qscore < 20. The depth reported reflects the number of acceptable basecalls, which can be different between the two programs.

Yes, it looks like Pisces and Mutect disagree. Pisces "0/." means it sees the reference allele but the second allele (the insertion) does not meet thresholds of confidence to call.

Sometimes looking at your BAMs in IGV or similar helps to see what is going on at the read level.

best Tamsen