Shiny-SoSV

Shiny-SoSV predicts somatic structural variant (SSV) detection performance for short-read whole genome sequencing experiments to inform study design.

Key Features:

• Parameter exploration: Investigates the impact of choice of variant detection tool, sequencing depth of coverage, variant allele fraction, and variant breakpoint resolution on SSV detection performance.
• Performance prediction: Models effects of single parameters and their combinations using simulation studies and a generalized additive model (GAM).
• Evaluation metrics: Reports sensitivity, precision, and F1 score to quantify SSV detection performance.
• Study design optimization: Enables prediction of how changes in experimental variables influence SSV detection capabilities to guide experimental planning.

Scientific Applications:

• Cancer genomics: Inform design and interpretation of studies of tumor development and evolution by predicting SSV detection under varying experimental conditions.
• Experimental design for short-read WGS: Guide selection of variant callers, sequencing depth, and target variant allele fractions to optimize detection sensitivity and precision.
• Clinical diagnostics: Support optimization of sequencing parameters to improve reliability and efficiency of somatic structural variant detection in diagnostic workflows.

Methodology:

Simulation studies of somatic structural variants followed by modeling detection performance using a generalized additive model (GAM) to assess the individual and combined effects of variant caller, sequencing depth, variant allele fraction, and breakpoint resolution on sensitivity, precision, and F1 score.