Computational aspects of DNA mixture analysis

TL;DR

This paper introduces a Bayesian network approach for efficient DNA mixture analysis, enabling local computations and improved inference in complex genetic data, along with diagnostic tools for model assessment.

Contribution

It presents a novel Bayesian network framework for DNA genotype analysis, enhancing computational efficiency and providing diagnostic tools for model validation.

Findings

Efficient likelihood computation using Bayesian networks.

Local computation reduces complexity of DNA mixture analysis.

Diagnostic tools improve model adequacy assessment.

Abstract

Statistical analysis of DNA mixtures is known to pose computational challenges due to the enormous state space of possible DNA profiles. We propose a Bayesian network representation for genotypes, allowing computations to be performed locally involving only a few alleles at each step. In addition, we describe a general method for computing the expectation of a product of discrete random variables using auxiliary variables and probability propagation in a Bayesian network, which in combination with the genotype network allows efficient computation of the likelihood function and various other quantities relevant to the inference. Lastly, we introduce a set of diagnostic tools for assessing the adequacy of the model for describing a particular dataset.