Multimodal representation learning for predicting molecule-disease relations

Jun Wen, Xiang Zhang, Everett Neil Rush, Vidul Ayakulangara Panickan, Xingyu Li, Tianrun Cai, Doudou Zhou, Yuk-Lam Ho, Lauren Costa, Edmon Begoli, Chuan Hong, J. Michael Gaziano, Kelly Cho, Junwei Lu, Katherine P. Liao, Marinka Zitnik, Tianxi Cai

Abstract: Predicting molecule–disease indications and side effects is important for drug development and pharmacovigilance. Comprehensively mining molecule–molecule, molecule–disease and disease–disease semantic dependencies can potentially improve prediction performance.