PIO

PIO optimizes selection of genomic intervals for next-generation sequencing (NGS) panels to maximize informativity (the proportion of patients with at least one detectable mutation) while minimizing panel length for cancer diagnostics.

Key Features:

• Optimization strategies: Supports defining genomic intervals at gene or exon level and optimizing selection based on patient-level informativity or informativity per kilobase of sequenced DNA.
• Custom panel enhancement: Suggests sets of genomic intervals to increase the informativity of existing custom panels without unnecessary panel expansion.
• Panel benchmarking: Provides a panel tester to benchmark custom and commercial NGS panels against informativity and panel size metrics.
• Efficiency improvements: Optimization algorithms can reduce panel size by up to 1000 kilobases while maintaining or improving mutation detection.
• Data sources: Operates on patient-level mutational data from private datasets or preloaded data from 91 independent cohorts covering 31 cancer types.
• Implementation: Implemented as an R package.

Scientific Applications:

• Cancer NGS panel design: Designing and optimizing cancer-focused NGS panels to maximize mutation detection while minimizing sequencing footprint.
• Panel benchmarking and comparison: Evaluating performance of custom and commercial NGS panels using informativity and panel size metrics.
• Cohort-driven panel tailoring: Tailoring panels to specific cancer types using patient-level mutational data from 91 cohorts across 31 cancer types.

Methodology:

Selection of genomic intervals (genes or exons) is optimized to maximize informativity and minimize panel length using either patient-level informativity or informativity per kilobase; analyses accept patient-level mutation data from private datasets or preloaded cohort data (91 cohorts, 31 cancer types) and include a panel tester function for benchmarking.