Systematic discovery of conservation states for single-nucleotide annotation of the human genome

Adriana Arneson,Jason Ernst

doi:10.1038/s42003-019-0488-1

Adriana Arneson, Jason Ernst

Open Access

https://doi.org/10.1038/s42003-019-0488-1

Copy DOI

Abstract

Comparative genomics sequence data is an important source of information for interpreting genomes. Genome-wide annotations based on this data have largely focused on univariate scores or binary elements of evolutionary constraint. Here we present a complementary whole genome annotation approach, ConsHMM, which applies a multivariate hidden Markov model to learn de novo ‘conservation states’ based on the combinatorial and spatial patterns of which species align to and match a reference genome in a multiple species DNA sequence alignment. We applied ConsHMM to a 100-way vertebrate sequence alignment to annotate the human genome at single nucleotide resolution into 100 conservation states. These states have distinct enrichments for other genomic information including gene annotations, chromatin states, repeat families, and bases prioritized by various variant prioritization scores. Constrained elements have distinct heritability partitioning enrichments depending on their conservation state assignment. ConsHMM conservation states are a resource for analyzing genomes and genetic variants.

Highlights

Comparative genomics sequence data is an important source of information for interpreting genomes
ConsHMM, to annotate a genome into conservation states at single nucleotide resolution based on a multiple species DNA sequence alignment (Fig. 1a, Methods)
In each state of the hidden Markov model (HMM), ConsHMM assumes that the probability of observing a specific combination of observations is determined by a product of independent multinomial random variables

Summary

Introduction

Comparative genomics sequence data is an important source of information for interpreting genomes. We applied ConsHMM to a 100-way vertebrate sequence alignment to annotate the human genome at single nucleotide resolution into 100 conservation states These states have distinct enrichments for other genomic information including gene annotations, chromatin states, repeat families, and bases prioritized by various variant prioritization scores. The representation of comparative genomics information into univariate scores or binary elements is limited in the amount of information it can convey about the underlying multiple sequence alignment at a specific base This limitation has become more pronounced given the large number of species available in multi-species alignments such as a 100-way alignment to the human genome[21]. An alternative approach learned patterns of different classes of mutations between human and only one non-human genome[29], and was only applicable at a broad region level

Methods

Results

Conclusion