Applications of probability and statistics in cancer genomics

Abstract

BackgroundThe past decade has witnessed a rapid progress in our understanding of the genetics of cancer and its progression. Probabilistic and statistical modeling played a pivotal role in the discovery of general patterns from cancer genomics datasets and continue to be of central importance for personalized medicine.ResultsIn this review we introduce cancer genomics from a probabilistic and statistical perspective. We start from (1) functional classification of genes into oncogenes and tumor suppressor genes, then (2) demonstrate the importance of comprehensive analysis of different mutation types for individual cancer genomes, followed by (3) tumor purity analysis, which in turn leads to (4) the concept of ploidy and clonality, that is next connected to (5) tumor evolution under treatment pressure, which yields insights into cancer drug resistance. We also discuss future challenges including the non‐coding genomic regions, integrative analysis of genomics and epigenomics, as well as early cancer detection.ConclusionWe believe probabilistic and statistical modeling will continue to play important roles for novel discoveries in the field of cancer genomics and personalized medicine.