A novel local manifold-ranking based K-NN for modeling the regression between bioactivity and molecular descriptors

• If the relationship between chemical structures and activity is overly complex, the global methods usually perform bad.
• As the structure of data usually shows complex, traditional methods can not detect the molecular similarities inherently.
• Manifold-ranking algorithm can detect the intrinsic similarity by graph.
• Thus, manifold ranking based K-NN is constructed for modeling the regression between bioactivity and molecular descriptors.

In the present study, we propose a novel local regression algorithm based on manifold-ranking and k-nearest neighbors (MRKNN for short). Under the framework of kernel methods, the group relationship shared among multiple molecules is firstly captured by the graph where nodes represent molecules and edges represent pairwise relations. Then, manifold ranking algorithm is developed for query-oriented extractive summarization, where the influence of query is propagated to other molecules through the structure of the constructed graph. When evaluated on four SAR datasets, MRKNN algorithm can provide a feasible way to exploit the intrinsic structure of similarity relationships. Results have validated the efficacy of the proposed algorithm.