Tore Eriksson

This study evaluates deep learning's role in improving quantitative phenotypes for GWAS in osteoporosis using bone X-ray images and UK Biobank data, finding limited improvements over traditional methods.

The paper uses latent representations of psychometric tests to detect genetic associations with dimensional autistic phenotypes, finding significant links in the domain of insistence on routine.

The paper uses UK Biobank data to predict protease networks linked to disease by analyzing genetic associations and phenome-wide results.