A novel and accelerated method for integrated alignment and variant calling from short and long reads

Area of Science:

• Genomics and Bioinformatics
• Computational Biology
• Molecular Biology

Background:

• Short-read sequencing (e.g., Illumina) offers high accuracy and cost-efficiency but struggles with structural variant (SV) detection and complex genomic regions.
• Long-read sequencing (e.g., PacBio HiFi) excels at resolving large SVs and repetitive sequences but has limitations in throughput, indel error rates, and cost.
• Hybrid approaches integrating both technologies leverage complementary strengths to enhance genomic analysis accuracy, comprehensiveness, and throughput.

Purpose of the Study:

• To benchmark the DNAscope Hybrid (DS-Hybrid) pipeline, a novel bioinformatics framework integrating short- and long-read sequencing data.
• To evaluate the DS-Hybrid pipeline's performance in detecting small variants (SNPs/indels), structural variants (SVs), and copy-number variations (CNVs).
• To compare the DS-Hybrid pipeline's efficiency and accuracy against existing tools like DeepVariant across human genome reference datasets.

Main Methods:

• The DNAscope Hybrid pipeline was developed as an integrated alignment and variant calling framework utilizing both short- (Illumina) and long-read (PacBio HiFi) data from the same samples.
• Performance was evaluated on human genome reference datasets (HG002-HG004) using Q100 and Genome in a Bottle v4.2.1 benchmarks.
• Variant calling capabilities (SNPs, indels, SVs, CNVs) were assessed at varying read depths (5×-30×) and compared to DeepVariant.

Main Results:

• The DS-Hybrid pipeline demonstrated significantly improved accuracy for SNP and indel calling, particularly in complex genomic regions.
• At lower long-read depths (5×-10×), the hybrid approach outperformed stand-alone pipelines at higher depths, reducing variant calling errors by over 50%.
• DS-Hybrid surpassed leading open-source tools in SV and CNV detection, showed clinical utility by identifying disease-associated variants, and achieved runtimes under 90 minutes.

Conclusions:

• The DNAscope Hybrid pipeline offers a computationally efficient and highly accurate solution for variant calling by integrating short- and long-read sequencing advantages.
• It enhances variant detection in challenging genomic regions, providing a robust framework for clinical applications and large-scale genomic studies.
• The pipeline shows significant promise for advancing genetic disease diagnostics, population-scale research, and personalized medicine initiatives.