Merve Nur Güler

The paper develops a multi-task neural network to predict blood metabolite profiles by separating covariate, genetic, and joint effects, finding nonlinear covariate modeling improves prediction.

This paper shows that ignoring gene-gene interactions in GWAS can lead to false positive results, highlighting the need for more complex models.

DeepKin predicts relatedness from low-coverage genomes and paleogenomes using CNNs, achieving accuracy >90% with 10K shared SNPs and outperforming READv2.

The paper analyzes genetic data from Neolithic Catalhöyük to show changes in kinship patterns and persistent female roles over time.

The paper identifies age-specific burial practices at Neolithic Catalhöyük by analyzing ancient DNA preservation differences between subadults and adults.

This study benchmarks four kinship estimation tools for ancient genomes using simulated pedigrees and finds varying accuracy depending on the tool and SNP count.

READv2 improves ancient DNA relatedness analysis by being faster, more accurate, and easier to use than previous methods.