Mutational Signature Analysis Segmentation

The patent segments the large set of mutational signatures into two distinct groups: a first set of common signatures and a second set of rare signatures. This segmentation allows the analysis to be divided into manageable stages - first fitting common signatures to establish a baseline, then selectively fitting rare signatures only when necessary to explain residual mutations. This resolves the contradiction by reducing computational complexity while maintaining comprehensive coverage through a systematic two-stage approach.

The patent performs preliminary action by first fitting the first set of common mutational signatures to the sample data before attempting to fit rare signatures. This preliminary fitting establishes a foundation that explains the majority of mutational patterns, and only the residual unexplained mutations are then subjected to analysis for rare signatures. This preliminary action reduces the computational burden while ensuring comprehensive analysis is not compromised.

By segmenting signatures into common and rare sets, the patent implements a time-efficient two-stage fitting process. The first stage quickly identifies common signatures that explain the majority of mutations, providing immediate accurate results. The second stage only processes rare signatures when needed, avoiding unnecessary computational time while maintaining high measurement precision for both common and rare signature identification.

The patent applies partial action by selectively fitting rare signatures only when the residual mutations after common signature fitting indicate their presence. This avoids the excessive action of fitting all rare signatures to every sample, which would waste time on samples where rare signatures are not relevant. The partial fitting approach maintains measurement precision by only analyzing rare signatures when they are actually needed to explain the data.