It then generates a diploid genome containing the sampled variants with an enhanced version of vcf2diploid (Rozowsky et al., 2011) (see Supplementary Material).
Pipeline commands are available in Additional file 5. Multi sample variant detection was performed with the same three software applications above as well as with the GATK HC (version 2.7-4-g6f46d11 [ 6, 7]) using default settings.
Single sample variant detection was performed with three different software applications: 1) SAM (MPileup / bcftools, version 0.1.18 [ 9]), 2) PL (version 0.5.2 [ 8]) and 3) the GATK UG (version 2.7-4-g6f46d11 [ 6, 7]) using default settings.
Cancer samples with variants with varying allele fractions are therefore particularly well suited to assess how well allele fractions are preserved during whole genome amplification.
Positions that have missing genotype information are ignored in the computation; therefore, the multisample VCF file should ideally consist of samples and variants with reasonably complete data.
The section is not organized in paragraphs, and mixes performance info (run time using different number of samples and variants) with example results.
Results I have two important criticisms about the Results section: The section is not organized in paragraphs, and mixes performance info (run time using different number of samples and variants) with example results.
Subsequent multi-sample variant calling with mpileup [ 48] yielded genotypes at 17.17 million sites.
However, with larger sample sizes, rare variants with potentially larger effect sizes and recessive modes of inheritance may be detected.
The results from our simulation studies indicate type I error rate is controlled, however, power falls quickly for small sample sizes using variants with modest effect sizes.
Hardy-Weinberg p-values were calculated using plink [ 27, 28] on a subset of 239 unrelated samples, and those variants with a p < 10-10 were removed.
