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.
A test of four genomic sequence-to-expression deep learning models (Enformer, Basenji2, ExPecto, Xpresso) finds that they often fail to predict the correct direction of effect of cis-regulatory genetic variation on gene expression.
The chemotherapeutic agent CX-5461 is shown to be a potent mutagen in hTERT-RPE1, HAP1 and human induced pluripotent stem cells. The compound generates distinct mutational patterns of single- and double-base substitutions, as well as of small insertions and deletions, that were detectable following a single exposure.
GLIMPSE2 is an improved method using sparse models for accurate, efficient and cost-effective genotype imputation in low-coverage whole-genome sequencing data.
Adjusting for common variant polygenic scores improves yield in gene-based rare variant association tests for quantitative traits, particularly when using sparse mixed models or simple linear models as an alternative to dense mixed-model approaches.