A good example of a test demonstrating the speedup is
'tests/sub1KG-tiny/EUR.via.filter.all22..impute.snps'. It requests 741
SNPs via --impute.snps and the software is otherwise given no 'hint'
as to the positions of the SNPs. Those SNPs are spread across all 22
chromosomes, but within just one window in each chromosome. Due to the
build database, the software quickly knows which windows to skip
straight to that are relevant for those 741 SNPs. You can see this in
the 'output.actual' file in that folder - you can see that only the
relevant windows are included.
test ssimp --gwas gwas/small.random.csv --ref ref/small.vcf.sample.vcf.gz --out output.txt --tag.snp listoftagsnps.txt
A good example of a test demonstrating the speedup is 'tests/sub1KG-tiny/EUR.via.filter.all22..impute.snps'. It requests 741 SNPs via --impute.snps and the software is otherwise given no 'hint' as to the positions of the SNPs. Those SNPs are spread across all 22 chromosomes, but within just one window in each chromosome. Due to the build database, the software quickly knows which windows to skip straight to that are relevant for those 741 SNPs. You can see this in the 'output.actual' file in that folder - you can see that only the relevant windows are included.
test ssimp --gwas gwas/small.random.csv --ref ref/small.vcf.sample.vcf.gz --out output.txt --tag.snp listoftagsnps.txt