Simulate meiosis from a panel of population vcfs (aka 1kG).
Recombination maps per chromosome need to be inputed. You can get recombination maps for Human (GRCh38) in the beagle docs.
The maps need to be in PLINK map format.
The vcf collections (aka 1 multisample-VCF per chromosome) are available from 1000genomes EBI ftp site.
The DECODE dataset contain +1k trios with an average DNM of 67 mutations/trio. They can be used (almost) directly with meiosim to introduce DNM by selecting one sample randomly, you can download this data from their paper.
Download, chr21 and chr22 from here. This will take a while.
Once this is done, you can downsample the vcfs to ~10Mb with:
vcfname="1kGP_high_coverage_Illumina.chr21.filtered.SNV_INDEL_SV_phased_panel.vcf.gz"
mkdir -p debug/vcfcollectionssmall
bcftools view -r chr21:0-10000000 -o debug/vcfcollectionssmall/chr21.vcf.gz ${vcfname}
Then to speed up the program you can also select the samples that you are going to use, it's also important to normalize the snps other-wise dwgsim will complain.
mkdir -p debug/vcfcollectionssmall2
files=$(ls debug/vcfcollectionssmall/*.vcf.gz)
for i in $files;
do
ibase=$(basename $i)
bcftools view -s NA21123,NA20752 $i | bcftools norm -m +snps | bcftools view -o debug/vcfcollectionssmall2/$ibase
bcftools index -t debug/vcfcollectionssmall2/$ibase
done
Download from beagle and change the chromosome names:
files=$(ls debug/recombmaps/*.map)
mkdir -p debug/recombmaps2
for i in $files;
do
ibase=$(basename $i)
awk '{print("chr"$0)}' $i > debug/recombmaps2/${ibase}
done
cd debug
wget https://static-content.springer.com/esm/art%3A10.1038%2Fnature24018/MediaObjects/41586_2017_BFnature24018_MOESM2_ESM.zip
unzip 41586_2017_BFnature24018_MOESM2_ESM.zip
cd nature24018-s2
tar -xvzf Aging_Oocytes_Supplementary_Table_DNMs.tar.gz
mv decode_DNMs/ ..
cd .. ## back to debug folder
cargo install --path .
meiosim \
-r debug/recombmaps2/ \
-v debug/vcfcollectionssmall2/ \
-d debug/decode_DNMs/vcfs/Proband-1354.vcf \
-p NA21123 \
-P NA20752 \
--prefix testout \
--seed 3 -f 5 \
--genome debug/hg38.genome
With the whole vcf the time is: 1m30s, with only the two parents, the time is: 10s
We also need to normalize the SNPs otherwise it's giving us problems down the road. This essentially means that we are removing multiallelic variants.
mutationfile=testout/sib0_NA21123_NA20752_meiosimvariants.txt
dwgsim -m ${mutationfile} -o 1 -c 0 -C 10 -R 0.0 smallgenome.fa outsib0