deploid-dev/deploid-bestpractices-github-actions

deploid-bestpractices-github-actions's Issues

sigma

use sigma 2.0

run script

#!/bin/bash
dataDir="~/dEploidPaper/deconv/"
root="/well/mcvean/joezhu/DEploid-Lasso-Benchmark/"
function run_dEploid {
mkdir ${root}"dEploidOut/"${sample}
cd ${root}"dEploidOut/"${sample}
#rm -r ErrFiles OutFiles
mkdir ErrFiles OutFiles

echo "
#!/bin/bash
#$ -cwd
#$ -V
#$ -P mcvean.prjc -q short.qc
#$ -e ErrFiles
#$ -o OutFiles
#$ -N ${sample}Ex
#$ -t 1-30

plaf=${dataDir}labStrains.eg.PLAF.txt
excludeAt=${dataDir}exclude.txt
vcf=${dataDir}${sample}.vcf.gz

panel=${root}combined_pf3k_field_panel_withcross_withparent.txt

for vqslod in $(seq 8 11); do
prefix=${sample}_fieldwithcross_withparent_vqslod${vqslod}_seed${SGE_TASK_ID}k$@
common="-vcf ${vcf} -plaf ${plaf} -exclude ${excludeAt} -o ${prefix}"
dEploidCommon="${common} -seed ${SGE_TASK_ID} -nSample 500 -rate 8 -burn 0.67 -sigma 1.6 -vqslod ${vqslod}"
rCommon="${common} -dEprefix ${prefix}"

time dEploid ${dEploidCommon} -panel ${panel} -ibd -k $@

&> ${root}/dEploidOut/${sample}/${prefix}.time

./interpretDEploid.r ${rCommon}
done

" > ${sample}_fieldk$@.sh
qsub ${sample}_fieldk$@.sh
}

while read sample ; do
run_dEploid 3
run_dEploid 4
done < labSampleNames

Check 401 seed 11 sigma 1.6

seed 1 - 15

vqslod

for i in $(seq 1 1 14); do bcftools view -f "PASS" --include 'VQSLOD &gt; 5' PG0389-C.wg.vcf.gz | grep -v "#" | grep Pf3D7_$(printf %02d $i)_v3 | wc -l; done

Run times

18000 sites

-noPanel 2 mins
Panel with size of 4, 3mins
-ibd 15 mins
All above times 15 times

-best only k

-best only p

-best
7 mins (sd. ) once