University of California, USA
Title: Discovery of Structural Modularity for Nanomedicine Design
Structural proteins of viruses have the capacity to function both in assembly and in disassembly, which is made possible through a built-in flexibility triggered by cellular events. Focusing on the highly selected viral capsid over their conformational domains between the metastable intermediates and the stable mature forms so far has provided us the essence of the needed stability, resisting extreme pH and digestive enzymes, to deliver medicals or agents to target tumors through both the circulation and the mucosal routes (1-3). Aided by multimodal imaging integrated with the deep learning via convolutional neural network to guide the tracking and the targeting of the payloads, the design principles of inserting heterologous epitopes to target the mucosal surfaces will be exemplified regarding in this presentation. Deep learning has gained enormous attention by the success of its convolutional neural networks in demonstrated machine learning tasks including high-content image classification. Crucial precision of cargo deliveries can be better realized through the AI deep-learning and the multiple modality of imaging domains to fully enable the targeting-engineered nanocapsids via various non-invasive mucosal routes. Further advancement of imaging technology like cryoEM and cellular tomography (see details in nobel.se; the Nobel Prize in Chemistry 2017) will be demonstrated in the unveiling of the associated molecular mechanisms essential to the platform vector design towards the success of constructing a non-invasive, mucosal targeting system (4-6).