Machine Learning and Deep Learning in Genetics and Genomics

In this chapter, we introduce various machine learning (ML) methods and deep learning (DL) algorithms, commonly adopted in genomics data analysis. We begin with a general introduction of genomics data and present a multi-omics study investigating early childhood oral health. We then review statistical methods and ML/DL methods and their application in genomics data analysis that include the following aspects: (1) association between genetic markers, mostly single nucleotide polymorphisms (SNPs), and complex diseases or traits in genome-wide association studies (GWAS), (2) copy number variation (CNV), and single nucleotide variant (SNV) calling in whole genome sequencing (WGS) or whole exome sequencing (WES) data of tumor samples, (3) association between DNA methylation status and phenotypes, which are commonly referred to as epigenome-wide association studies (EWAS), (4) analysis of genome-wide high-throughput chromosome conformation capture (Hi-C) data, (5) inference related to transcription factor binding sites (TF), and (6) single-cell RNA-seq data analysis. To complete the review, we present the results of a systematic review of the machine learning landscape in oral diseases. We conclude with a discussion of potential future applications of ML/DL in genetics and genomics in oral health.