Fig. 8
From: Harnessing large language models (LLMs) for candidate gene prioritization and selection
Overview of LLM-generated scores for A37 module genes. This network represents the scores generated by two LLMs, GPT-4 and Claude, for six prespecified criteria, across a pool of candidate genes distributed across 11 BloodGen3 modules from the aggregate A37. Each pie chart shows the relative magnitude of the averaged score from the two models. The size of each chart is proportional to the overall cumulative score across the six criteria