deconstructSigs

deconstructSigs decomposes somatic mutation catalogs from individual tumor samples into weighted combinations of published mutational signatures to identify mutational processes shaping cancer genomes.

Key Features:

• Input Data: Uses per-sample fractions of somatic mutations across the 96 trinucleotide contexts as the input mutation spectrum.
• Signature Identification: Fits a weighted combination of published mutational signatures to reconstruct the observed mutational profile of a single tumor sample.
• Analytical Capability: Discerns distinct and dynamic mutational processes that can vary between cancer types, for example esophageal adenocarcinoma and squamous cell carcinomas.

Scientific Applications:

• DNA Repair Deficiencies: Identifies samples with DNA repair deficiencies by detecting associated mutational signatures.
• Environmental Exposures: Attributes mutational processes driven by environmental factors through their characteristic signatures.
• Precision Medicine: Reveals tumor-specific mutagenic processes to inform precision cancer medicine and guide tailored therapeutic strategies.

Methodology:

Fits weighted combinations of published mutational signatures to sample-specific fractions of mutations across the 96 trinucleotide contexts to reconstruct each tumor's mutational profile.