September 26, 2024
The Center for Translational Bioinformatics (CTBI) at WashU Medicine is an interdisciplinary initiative, integrating sophisticated computational methodologies with translational research to drive advancements in precision medicine. Embedded within the Institute for Informatics, Data Sciences, and Biostatistics and the McDonnell Genome Institute, CTBI serves as a unified platform for transforming complex, multi-omic data into actionable clinical insights.
Mission and Structure
CTBI’s mission is to address key translational barriers by organizing its efforts into three core pillars:
- Translational Research: CTBI develops advanced bioinformatics frameworks to facilitate the integration of diverse data modalities—genomics, proteomics, and imaging—enabling more efficient linkage of molecular data to clinical phenotypes.
- Precision Therapeutics: The center focuses on constructing systems that support the precise matching of therapeutic interventions to individual patient profiles, leveraging multi-omic data to inform data-driven, personalized treatment strategies.
- Healthy Aging: In its second year, CTBI will investigate the molecular underpinnings of aging, with the goal of identifying biomarkers for longevity and disease prevention, ultimately shaping precision-based approaches tailored to the genetic and environmental factors influencing the aging process.
By creating a unified framework that integrates multimodal data—from genomics to clinical phenotypes—we are enabling a new frontier in precision medicine where patient care is informed by comprehensive, data-driven insights at the molecular and systems levels.
Rich Head, Inaugural Director of the Center of Translational BioInformatcs (CTBI)
Overcoming Key Challenges
CTBI is designed to address three critical challenges in advancing precision medicine:
- Siloed Data Systems: A major barrier in translational research is the isolation of clinical and research data. CTBI is developing secure data integration systems that link clinical and research datasets while preserving patient confidentiality. A translational identifier system allows researchers to trace clinical samples back to their origin, enabling the use of clinical data in research without compromising privacy.
- Data Complexity and Scale: WashU Medicine generates vast, multimodal datasets from genomics, metabolomics, imaging, and clinical records. CTBI’s advanced computational infrastructure, enhanced by AI, manages and analyzes these complex datasets, extracting clinically relevant insights at scale.
- Cross-Disciplinary Integration: CTBI bridges bioinformatics, clinical informatics, machine learning, and computational biology, fostering collaboration between researchers, clinicians, and data scientists. This interdisciplinary approach ensures that computational tools are effective and widely applicable, accelerating the translation of research into clinical practice.
Mixed Method AI: A Hybrid Approach to Precision Medicine
CTBI’s computational strategy leverages Mixed Method AI, integrating model-based and knowledge-based AI to analyze complex, multimodal datasets at scale while maintaining biological context.
- Model-Based AI: Utilizing machine learning, model-based AI excels at detecting patterns in large, structured datasets like genomic and proteomic profiles, enabling rapid, high-throughput analysis. However, its limitations emerge in novel scenarios where data deviate from the training set, reducing its capacity for generating new hypotheses or interpreting complex biological relationships.
- Knowledge-Based AI: In contrast, knowledge-based AI mimics human reasoning, contextualizing data within biological frameworks. It can infer relationships and generate novel hypotheses by integrating multiple data types, offering mechanistic insights not reliant on predefined training sets.
- Combined Approach: CTBI’s hybrid system merges these methods, enabling efficient pattern recognition with model-based AI and deep, interpretative reasoning through knowledge-based AI. This synergy drives not only data analysis but also hypothesis generation, experimental design, and therapeutic discovery.
Scaling the System: Early Success and Future Directions
CTBI’s computational framework has already demonstrated its value in multiple publications that highlight the system’s ability to derive novel insights from complex datasets. Two key publications exemplify this approach:
NAD Precursors and Bile Acid Sequestration for Treating Pre-Clinical Environmental Enteric Dysfunction
Published in Science of Translational Medicine, this study examined environmental enteric dysfunction (EED) in children from Zambia and Pakistan. CTBI’s AI systems analyzed RNA sequencing data to identify key biological processes, validated in animal models. Supplementation with NAD precursors and bile acid sequestration, informed by the AI analysis, successfully reversed physiological damage in preclinical models, presenting potential new treatments for EED in children.
Molecular Subtypes of Crohn’s Disease and Clinical Outcomes
Published in Gastroenterology, this study used CTBI’s Mixed Method AI to identify five molecular subtypes of adult Crohn’s disease through RNA sequencing of surgical resections. By linking these subtypes to clinical outcomes, including post-surgery recurrence rates, the analysis revealed crucial insights into disease progression and therapeutic targets. This integrative approach enhances understanding of Crohn’s disease biology and supports personalized treatment strategies.
Scaling to Large Disease Areas
CTBI is expanding its platform to address larger disease areas, starting with Alzheimer’s disease. By integrating human omics and animal model data, CTBI aims to map disease pathways, identify novel therapeutic targets, and discover predictive biomarkers using its Mixed Method AI. The platform is designed to scale for other neurodegenerative diseases, like ALS and Parkinson’s, and additional complex diseases, bridging preclinical research with human studies at an unprecedented scale.
Future Directions: Building for Impact
CTBI is positioned as a central player in WashU Medicine’s precision medicine initiative. By continuously expanding the scope of its computational platforms, CTBI ensures that cutting-edge AI techniques are applied to the most pressing challenges in translational research. The ultimate goal is to create a dynamic, data-driven ecosystem where computational tools directly inform patient care, leading to more precise diagnoses, personalized therapies, and improved clinical outcomes.
Learn more about CTBI from our recent inaugural seminar:
Are you a WashU Medicine research who would like to collaborate?
