Ai+ Training

 In a global pandemic such as COVID-19, technology, artificial intelligence, and data science have become critical to helping societies effectively deal with the outbreak. In this talk, I will discuss three case studies of how AI is being used in Corona Virus research. The first part of the talk will discuss how the deep learning model detected COVID-19 caused pneumonia from computed tomography (CT) scans with comparable performance to expert radiologists. To be more specific, I will discuss the UNet++ architecture that was implemented by researchers for evaluating lung infection in COVID-19 CT images. The second part of the talk will be devoted to recent attempts in natural language processing to generate new insights in support of the ongoing fight against this infectious disease. There is a growing urgency for these approaches because of the rapid acceleration in new coronavirus literature, making it difficult for the medical research community to keep up. To be precise, BERT literature search engine for COVID-19 literature. will be discussed.

The third part of the talk deals with deep learning-based generative modeling framework to design drug candidates specific to a given target protein sequence. One of the most important COVID-19 protein targets is the 3C-like protease for which the crystal structure is known. We present different deep learning models designed for generating novel drug molecules with multiple desirable properties. The deep learning framework involves Variational Autoencoder, Generative Adversarial Networks, Reinforcement Learning, and Transfer Learning. The generated molecules might serve as a blueprint for creating drugs that can potentially bind to the viral protein with high target affinity, as well as high drug-likeliness. Last but not the least, this talk will also touch upon how the world community responded by making the data available to the researchers which enabled the data scientists to explore and support the scientific community.