Josip Rudar

This paper evaluates H5Nx influenza virus genotyping tools and finds they often group viruses with distinct evolutionary histories due to reassortment and reference bias.

This paper introduces WaveSeekerNet, a calibrated AI model that predicts IAV host origin and identifies key adaptive mutations, revealing host-adaptive distance and the Mammalian Adaptation Zone.

WaveSeekerNet predicts IAV subtypes and host source using attention-based deep learning, outperforming existing models in accuracy and efficiency.

The paper uses decision tree ensembles with multivariate splits to analyze 16S amplicon sequencing data, improving the analysis of microbial community differences in Crohn's Disease and colorectal cancer patients.

This study shows that COI primer choice affects arthropod richness and indicator taxa recovery in freshwater samples, suggesting multiple primer sets should be used for maximum diversity detection.