Missouri Data Science & Informatics Symposium 2024

Title: DNA -based data storage

Bio
Dr. de Figueiredo is a NextGen Precision Health Endowed Professor of Molecular Microbiology & Immunology. He has broad experience and training in molecular biology, cell biology, microbiology, and biochemistry, with specific expertise in host-pathogen interactions and biotechnology development. He has co-founded several biotechnology companies, including HelioWave Technologies, LLC, and authored over 70 peer-reviewed publications. His research has been supported by awards from the Bill & Melinda Gates Foundation, ARPA-h, NIH, NSF, USDA, DTRA, and DARPA, as well as several other private foundations.  

Title: Large Protein Language Models and Their Prompt-based Learning

Abstract
Protein language models (PLMs) provide a powerful representation of protein sequences and their evolutions through pre-training on vast protein sequence datasets. We introduced a structure-aware PLM, S-PLM, by integrating sequence and structure information to improve protein prediction. The model utilizes a multi-view contrastive learning strategy to align protein sequences with their structures within a shared embedding space. S-PLM leverages Swin-Transformer on the contact map images of AlphaFold-predicted structures fused with sequence-based embeddings from ESM2. It has a comprehensive set of fine-tuning tools that increase its prediction capacity, surpassing other PLMs. Additionally, we developed Prot2Seq to expand PLMs’ capabilities for multitasking protein predictions, utilizing an autoregressive language modeling method. By adding task-specific tokens into the decoder, Prot2Seq showed enhanced performance when conducting simultaneous multiple-task training within a single model run. Furthermore, we implemented a Parameter-Efficient Fine-Tuning framework, PEFT-SP, with various prompting methods, like Prompt Tuning and Adapter Tuning, on the ESM-2 model for predicting signal peptides. PEFT-SP gained significant prediction accuracy for signal peptides over other methods, especially when having limited training data. Our studies show great promise for PLMs and prompt-based learning in protein prediction tasks.

Bio
Dong Xu is Curators’ Distinguished Professor in the Department of Electrical Engineering and Computer Science, with appointments in the Christopher S. Bond Life Sciences Center and the Informatics Institute at the University of Missouri-Columbia. He obtained his Ph.D. from the University of Illinois, Urbana-Champaign in 1995 and did two years of postdoctoral work at the US National Cancer Institute. He was a Staff Scientist at Oak Ridge National Laboratory until 2003 before joining the University of Missouri, where he served as Department Chair of Computer Science during 2007-2016 and Director of Information Technology Program during 2017-2020. Over the past 30 years, he has conducted research in many areas of computational biology and bioinformatics, including single-cell data analysis, protein structure prediction and modeling, protein post-translational modifications, protein localization prediction, computational systems biology, biological information systems, and bioinformatics applications in human, microbes, and plants. His research since 2012 has focused on the interface between bioinformatics and deep learning. He has published more than 400 papers with more than 21,000 citations and an H-index of 73 according to Google Scholar. He was elected to the rank of American Association for the Advancement of Science (AAAS) Fellow in 2015 and American Institute for Medical and Biological Engineering (AIMBE) Fellow in 2020.

Theme: Artificial Intelligence for Good: An Industry Perspective

Title: The Evolving Role of AI in Healthcare: Opportunities and Real-World Challenges

Abstract: The integration of Artificial Intelligence (AI) in healthcare is transforming the delivery of care, offering unprecedented opportunities to enhance patient care, diagnostics, and operational efficiency. However, alongside its benefits, AI introduces real-world challenges, including ethical considerations, data privacy issues, and the complexities of integrating AI into existing clinical workflows. This keynote will explore the evolving role of AI in healthcare, showcasing its potential to revolutionize care delivery while addressing the obstacles that must be overcome to fully realize its promise in real-world applications.

Bio
Dr. Saif Khairat is the Beerstecher-Blackwell Distinguished Term Scholar and professor at the University of North Carolina at Chapel Hill (UNC). Dr. Khairat holds joint appointments at the Lineberger Comprehensive Cancer Center, Cecil G. Sheps Center for Health Services Research, and the School of Nursing. He is the associate director of the Carolina Health Informatics Program and the Principal Investigator (PI) of a five-year NIH-funded Center for Virtual Care Value and Equity (ViVE).

Dr. Khairat has authored over 100 influential scientific articles in over 35 different peer-reviewed journals and conferences. Moreover, he has been an investigator for over $7.5 million in research grants over the past five years.
Internationally, Dr. Khairat provides expert opinions to the World Health Organization (WHO) to inform national policies and strategies for digital health targeting non-communicable diseases in low- and middle-income countries.

Dr. Khairat earned his PhD in Health Informatics, bachelor’s and master’s in Computer Science from the University of Missouri, and an MPH in Health Policy and Management from the Gilling’s School of Global Health at the University of North Carolina at Chapel Hill.

Title: Enhancing Comparative Oncology with Transfer Learning of Gene Expression Programs

Abstract: 
Immunotherapy is a highly promising approach to treating cancer that harnesses the body's innate and adaptive defense mechanisms to suppress tumors for a longer duration compared to chemotherapy treatments. However, some types of cancer can avoid detection by the immune system and continue to develop. Optimizing cancer immunotherapy strategies requires a thorough understanding of the various cells and their states within the tumor microenvironment (TME), particularly their roles in immune evasion. Dogs provide an excellent model for investigating TME molecular processes due to their spontaneous development of cancers with tumor progression from primary to metastasis, which closely mirrors that in humans. However, our genomic understanding of dog TMEs is limited by a lack of species-specific reagents and datasets.
Single-cell RNA sequencing (scRNA-seq) has revolutionized the analysis of the human TME, highlighting the potential benefits of projecting molecular mechanisms learned from human data onto canine data. Yet, current cross-species analyses often rely on data integration that primarily focuses on shared cell types and introduces technical challenges such as batch effects. To address these issues, we propose a framework to identify Gene Expression Programs associated with cell types and biological processes in humans and transfer them across species using a transfer learning approach.
We expect our computational experiments to build the foundations of a comparative framework toward understanding, more broadly, immune suppression mechanisms in human and canine cancers, and facilitate better therapeutic outcomes in canine patients undergoing immunotherapy treatments.

Title:  AI for Good: A call to action

Abstract:

Bio:
Keith Wilson is an experienced AI leader specializing in delivering advanced AI solutions in the public sector. As a Senior Manager at C3 AI, he leads teams developing cutting-edge AI applications that address critical challenges such as contested logistics and generative AI implementation. A former U.S. Navy Nuclear Submarine Officer, Keith combines his strategic military background with expertise in AI to drive digital transformation and enhance operational efficiency. He is passionate about responsible AI, mentoring future leaders, and contributing to national security through innovative technology."