Ardan Yılmaz

The paper presents a structure-agnostic framework for predicting drug-drug interactions using MeSH semantics and a Drug-Target-Protein knowledge graph, showing improved performance with intermediate MeSH annotation depth.

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

This study introduces PROBE, a benchmarking framework for evaluating protein representations on function-related tasks, demonstrating its use with ESM2 and three multimodal PLMs.

MTaxi classifies closely related species from low coverage ancient DNA data using mitochondrial DNA substitutions, achieving 100% accuracy on tested samples.