paperKB
coga / coga-kb
Help
Sign in

Chunk #16 — 2 METHODS — 2.1 SNVMix model specification

Source
SNVMix: predicting single nucleotide variants from next-generation sequencing of tumors.
Embedded
yes

Text

Our approach to inference involves learning the parameters θ by fitting the model to training data using MAP EM (see below). We demonstrate that this produces better results than Maq, which uses fixed parameters (Section 3).