SigsPack

SigsPack estimates exposure to known mutational signatures for single samples using quadratic programming algorithms.

Key Features:

• Single-sample signature estimation: Exposures to known mutational signatures are estimated for individual samples rather than cohorts.
• Quadratic programming: Signature exposures are computed using quadratic programming algorithms.
• Stability assessment via bootstrapping: Bootstrapping input mutational catalogues provides estimations of exposure stability.
• Sequence composition normalization: The effect of sequence composition of mutational context can be accounted for by normalising the catalogues.

Scientific Applications:

• Mutational signature analysis: Quantification of known mutational signature exposures in individual tumor samples.
• Exposure stability evaluation: Assessment of confidence in signature exposure estimates through bootstrap resampling.
• Context-aware signature deconvolution: Normalization for mutational context sequence composition to improve exposure accuracy.

Methodology:

SigsPack employs quadratic programming algorithms to estimate mutational signature exposures from single-sample mutational catalogues, with bootstrapping to assess stability and optional normalization for sequence composition of mutational context.