Jul 2026
MEGA-CL: A Molecular Foundation Model for Generalizable ADMET Prediction through Graph External Attention and Contrastive Learning
The potential of MEGA-CL as a generalizable framework for accelerating in silico ADMET evaluation and early-stage drug candidate optimization is demonstrated, with robust performance on challenging regression tasks, including clearance (CL) and steady-state volume of distribution (VDss).
Ting-Hui Jin, Ke-Du Jin, Ying Li et al.
· arXiv.org · 0 citations