Thank you for visiting nature.com. You are using a browser version with limited support for CSS. To obtain
the best experience, we recommend you use a more up to date browser (or turn off compatibility mode in
Internet Explorer). In the meantime, to ensure continued support, we are displaying the site without styles
and JavaScript.
Machine learning methods are becoming increasingly important in the analysis of large-scale genomic, epigenomic, proteomic and metabolic data sets. In this Review, the authors consider the applications of supervised, semi-supervised and unsupervised machine learning methods to genetic and genomic studies. They provide general guidelines for the selection and application of algorithms that are best suited to particular study designs.
A large proportion of genetic variants in the human genome have been predicted to be deleterious. This Review examines the frequency and patterns of deleterious alleles in the human genome and considers recent studies with conflicting findings on whether the mutation load, or burden of deleterious alleles, differs across populations.
High-throughput DNA sequencing technologies are providing an ever-expanding wealth of genome sequence data, including detailed information on human genetic variation. However, such data typically lack haplotype information (that is, thecis-connectivity of variants along individual chromosomes). This Review describes diverse recent experimental methods by which genetic variants can be resolved into haplotypes, accompanying computational methods and important applications of these methods in genomics and biomedical science.
The analysis of whole-genome sequence data from both modern and ancient humans has provided evidence for archaic adaptive introgression. Here, the authors provide an overview of the statistical methods used and the supporting evidence for reported examples of archaic introgression, which may have driven the acquisition of beneficial variants that enabled adaptation and survival in new environments.
