Kim Minseo

This paper identifies issues with the RELAX software's reproducibility and reliability, showing that it fails to reproduce its own estimates and outputs inconsistent results.

The paper tests how branch-labeling methods detect selection on real genes versus random labels, finding high rejection rates even when labels are random, suggesting the methods are sensitive to branch-specific variation in selection parameters.

The paper evaluates DNA foundation models for chimera detection in long-read metagenomic bins, finding that ensembling improves precision over single models.

This paper quantifies how Evo2 likelihood scores calibrate with fitness across five E. coli datasets, showing context-dependent calibration that varies by selection regime and sequence divergence.