A flaw in using pretrained protein language models in protein-protein interaction inference models
• Abstract With the growing pervasiveness of pretrained protein language models (pLMs), pLM-based methods are increasingly being put forward for the protein-protein interaction (PP