Neural networks for genetic epidemiology: past, present, and future

Abstract

During the past two decades, the field of human genetics has experienced an information explosion. The completion of the human genome project and the development of high throughput SNP technologies have created a wealth of data; however, the analysis and interpretation of these data have created a research bottleneck. While technology facilitates the measurement of hundreds or thousands of genes, statistical and computational methodologies are lacking for the analysis of these data. New statistical methods and variable selection strategies must be explored for identifying disease susceptibility genes for common, complex diseases. Neural networks (NN) are a class of pattern recognition methods that have been successfully implemented for data mining and prediction in a variety of fields. The application of NN for statistical genetics studies is an active area of research. Neural networks have been applied in both linkage and association analysis for the identification of disease susceptibility genes.In the current review, we consider how NN have been used for both linkage and association analyses in genetic epidemiology. We discuss both the successes of these initial NN applications, and the questions that arose during the previous studies. Finally, we introduce evolutionary computing strategies, Genetic Programming Neural Networks (GPNN) and Grammatical Evolution Neural Networks (GENN), for using NN in association studies of complex human diseases that address some of the caveats illuminated by previous work.

Grammatical Evolution Decision Trees for Detecting Gene-Gene Interactions

“…To overcome this problem, we propose using grammatical evolution to build "white box" models that are readily, immediately understandable. Similar approaches have been successful in other fields [14][15][16], strengthening the hypothesis that this approach would be successful in human genetics. Additionally, similar machine learning approaches have been shown to be successful in genetic applications.…”

Initialization parameter sweep in ATHENA

“…Even at the smallest population size of 50, where the deme size would have been very small (approximately 12), parallel ATHENA still achieved greater detection power. This serves as additional evidence for the benefit of the “island model” and suggests that migration of best solutions between demes prevents stalling on local minima in the fitness landscape [11]. …”

“…A periodic exchange or “migration” of best solutions between demes occurs a set number of times during each run. This exchange increases diversity among the solutions in the different populations and decreases stalling at local minima [11]. We wanted to determine if there was a threshold at which dividing the population size into smaller subpopulations was no longer beneficial.…”

Athena: A Tool for Meta-Dimensional Analysis Applied to Genotypes and Gene Expression Data to Predict HDL Cholesterol Levels

“…GENN uses a variation of genetic programming (GP) called grammatical evolution (GE) to optimize artificial neural networks to identify a model that predicts a given outcome 21–23 . GP is a computational technique that uses concepts of survival of the fittest in order to evolve a fit solution from an original population of random solutions 24 .…”