LPRC: Manifold Learning of Quantum Information for Drug Development DUIRI - Discovery Undergraduate Interdisciplinary Research Internship Summer 2024 Accepted Machine Learning; Computational Chemistry; Drug Discovery and Development Almost every biological event centers around intermolecular interactions, upon which drug molecules act to achieve inhibitory or promoting effects on the biological machinery of DNAs, RNAs, proteins, cytokines, and cells. The essence of drug discovery and development, thus, relies on searching for a drug molecule that can achieve a specific intermolecular interaction activity with desired biomolecules. How to predict a molecule's capability or specificity to interact with another molecule has been a fest, as well as a struggle, in the in silico design of drug products. We have developed novel machine-learning methods, including unsupervised and supervised, to encode quantum information of a molecule for the de novo design of drug molecules and supramolecular structures. The concept lies in manifold embedding and manifold learning, and using artificial neural networks as the universal approximator to bridge molecular structure and molecular functionality. Our concept allows both predictive and generative AI models to be built, demonstrating paradigm-shifting potentials in our preliminary studies. We are barely scratching the surface of chemical learning with quantum information and actively extending our research activities. Tonglei Li Tonglei Li The proposed work includes literature search, data curation, python programming, and running machine learning programs on CPU and GPU clusters. Students are expected to collect computational results, write reports for publication, and present the results. Students should have solid backgrounds in physical (organic) chemistry and (statistical) mathematics. Fair experiences with coding (Python, C, etc.) are desired; shell scripting is a plus. 3 40 (estimated)
This project is not currently accepting applications.