Search

Abstract Background Accurate calling of SNPs and genotypes from next-generation sequencing data is an essential prerequisite for most human genetics studies. A number of computational steps are required or recommended when translating the raw sequencing data into the final calls. However, whether each step does contribute to the performance of variant calling and how it affects the accuracy still remain unclear, making it difficult to select and arrange appropriate steps to derive high quality variants from different sequencing data. In this study, we made a systematic assessment of the relative contribution of each step to the accuracy of variant calling from Illumina DNA sequencing data. Results We found that the read preprocessing step did not improve the accuracy of variant calling, contrary to the general expectation. Although trimming off low-quality tails helped align more reads, it introduced lots of false positives. The ability of markup duplication, local realignment and recalibration, to help eliminate false positive variants depended on the sequencing depth. Rearranging these steps did not affect the results. The relative performance of three popular multi-sample SNP callers, SAMtools, GATK, and GlfMultiples, also varied with the sequencing depth. Conclusions Our findings clarify the necessity and effectiveness of computational steps for improving the accuracy of SNP and genotype calls from Illumina sequencing data and can serve as a general guideline for choosing SNP calling strategies for data with different coverage.

Abstract Background When using Illumina high throughput short read data, sometimes the genotype inferred from the positive strand and negative strand are significantly different, with one homozygous and the other heterozygous. This phenomenon is known as strand bias. In this study, we used Illumina short-read sequencing data to evaluate the effect of strand bias on genotyping quality, and to explore the possible causes of strand bias. Result We collected 22 breast cancer samples from 22 patients and sequenced their exome using the Illumina GAIIx machine. By comparing the consistency between the genotypes inferred from this sequencing data with the genotypes inferred from SNP chip data, we found that, when using sequencing data, SNPs with extreme strand bias did not have significantly lower consistency rates compared to SNPs with low or no strand bias. However, this result may be limited by the small subset of SNPs present in both the exome sequencing and the SNP chip data. We further compared the transition and transversion ratio and the number of novel non-synonymous SNPs between the SNPs with low or no strand bias and those with extreme strand bias, and found that SNPs with low or no strand bias have better overall quality. We also discovered that the strand bias occurs randomly at genomic positions across these samples, and observed no consistent pattern of strand bias location across samples. By comparing results from two different aligners, BWA and Bowtie, we found very consistent strand bias patterns. Thus strand bias is unlikely to be caused by alignment artifacts. We successfully replicated our results using two additional independent datasets with different capturing methods and Illumina sequencers. Conclusion Extreme strand bias indicates a potential high false-positive rate for SNPs.