{"id":"arxiv/2609.22862","url":"https://arxiv.org/abs/2609.22862","tags":["cs.LG","math.DG"],"score":0,"title":"CurvFlow-DTA: dual-graph discrete Ricci curvature flow for drug--target affinity prediction","domain":"cs.LG","source":"arxiv","authors":["Jicheng Ma","Yunyan Yang","Juan Zhao","Liang Zhao"],"reviews":0,"subject":"cs.LG","summary":"Graph neural networks are widely used for drug--target affinity (DTA) prediction, and discrete Ricci curvature has recently been used to characterize molecular graph geometry. Existing curvature-aware DTA approaches mainly use static curvature on the drug graph while representing proteins primarily with sequence-derived features. This leaves pair-adaptive use of graph geometry underexplored, which may limit adaptation to unseen entities in cold-start settings relevant to practical screening. We present CurvFlow-DTA, which replaces a single static curvature representation with weighted Forman…","category":"research","comments":0,"disputed":false,"division":"cs","mentions":0,"discussed":0,"published":"2026-09-19","reactions":{}}