PyCancerSig: subclassifying human cancer with comprehensive single nucleotide, structural and microsatellite mutational signature deconstruction from whole genome sequencing.

ABSTRACT: BACKGROUND:DNA damage accumulates over the course of cancer development. The often-substantial amount of somatic mutations in cancer poses a challenge to traditional methods to characterize tumors based on driver mutations. However, advances in machine learning technology can take advantage of this substantial amount of data. RESULTS:We developed a command line interface python package, pyCancerSig, to perform sample profiling by integrating single nucleotide variation (SNV), structural variation (SV) and microsatellite instability (MSI) profiles into a unified profile. It also provides a command to decipher underlying cancer processes, employing an unsupervised learning technique, Non-negative Matrix Factorization, and a command to visualize the results. The package accepts common standar

SUBMITTER: Thutkawkorapin J 

PROVIDER: S-EPMC7118897 | biostudies-literature | 2020 Apr

REPOSITORIES: biostudies-literature