SeqMule

SeqMule automates variant calling and consensus generation from next-generation sequencing (NGS) data for human genome and exome analysis.

Key Features:

• Integration of Multiple Tools: Incorporates five alignment tools and five variant calling algorithms, enabling analyses using multiple aligner–caller combinations.
• Cluster-Free Parallelization: Implements parallelization that does not require computational clusters and supports Sun Grid Engine and deployment on Amazon Web Services for distributed processing.
• Consensus Variant Calling: Normalizes and intersects variant calls from multiple callers to produce a consensus set, reducing discrepancies among aligners and callers with reported lower Mendelian error rates and increased consistency versus single-tool outputs.
• Annotated VCF Output and Performance: Produces annotated Variant Call Format (VCF) files and can generate results from a 30X whole-genome sequencing dataset within a day on example hardware (2 Intel Xeon X5650 CPUs, 48 GB memory) when fast turnaround is prioritized.
• Quality and Consistency Evaluation: Performs quality checks, Mendelian error checking, consistency evaluation, and generates HTML-based reports for result inspection.

Scientific Applications:

• Mendelian disease variant identification: Enables identification of disease-contributory variants from NGS data in studies of Mendelian disorders.
• Genetic research and clinical diagnostics: Supports genetic research and clinical diagnostic workflows by providing automated, multi-caller variant call sets for downstream interpretation.

Methodology:

Uses five alignment tools and five variant calling algorithms, performs normalization and intersection of variant calls to produce consensus VCFs, includes cluster-free parallelization (Sun Grid Engine, AWS), and conducts quality checks and Mendelian error checking.