Division of Cardiovascular Diseases Strategic Plan

Goals in Enabling Technologies and Methodologies for Cardiovascular Disease

1.3. Develop dynamic, predictive computational models of cell/organelle function in health and disease

Table of Contents

• Overview
• Strategies to Accomplish This Goal May Entail
  • Basic Research
  • Translational Research
  • Clinical Research
• Contributing Sources

Return to Goals in Enabling Technologies and Methodologies for Cardiovascular Diseases
Return to Goals of the Division of Cardiovascular Diseases Strategic Plan

Overview

A major roadblock to determining the causes and progression of disease is the complexity of biological systems. This complexity includes genetic and epigenetic variation, gene-gene and gene-environment interactions, and environmental factors (including both internal and external factors). Dynamic computational models are essential for understanding the complex factors underlying disease and for driving discovery and in silico testing of new therapeutic strategies. Barriers have included lack of tools and technologies for dynamic subcellular measurements of key molecules, inability to effectively share data and models, and a need for focus and cohesion within the scientific community.

Strategies to Accomplish this Goal May Entail:

Basic Research:

• Develop new approaches to integrate diverse data types into multi-scale predictive models. Data sources could include imaging, ‘omics, genetics, and other assays of cell function.
  • Understand determinants of heart development by focusing on cardiomyocyte differentiation and function at different developmental stages, including signaling and regulatory pathways, genetic and epigenetic variation, and physical and mechanical factors.
  • Develop models that are specialized for particular cell types or organelles to support specific research communities and their research focus (e.g., atherosclerosis, heart failure, ischemia, arrhythmia). These models should be generated and used broadly within targeted research communities.
    • Integrate knowledge from the mitochondrial proteome and genetic variation (in nuclear and mitochondrial DNA) to predict and understand mitochondrial phenotype and functioning during normal, acute and chronic ischemic conditions and in the cardio-protected state.
    • Encourage multi-disciplinary teams to ensure appropriate biological expertise.
• Improve data collection to enhance accuracy of computational models.
  • Develop new tools and technologies to measure subcellular localization and concentration of key molecules within cell types/organelles of interest.
  • Acquire sufficiently granular data to build fully specified, predictive, computational models, especially dynamic (temporal and circadian) and/or in vivo measurements at subcellular resolution.

Translational Research:

• Ensure model usability by researchers across computational, basic, and clinical realms. 
  • Construct models that are accessible to the research community, such that a CV researcher could compare her/his experimental results with model predictions, as well as with data that were used to build and validate the model.
  • Create predictive models that generate testable hypotheses of underlying function and/or responses to treatments. For instance, modeling the specific biological pathways in the pathophysiological chain of atherothrombosis for the ultimate purpose of intervention discovery.
  • Include clinically appropriate model design and study results in the development, validation, and use of models to ensure that these models have translational and clinical relevance. An example is the use of computational models of heart development regulatory networks to prioritize candidate genes in lieu of validation in large population datasets.

Clinical Research:

• Determine how changes in genome sequence affect protein structure, function, concentration, and/or localization. Knowledge-based predictive and personalized medicine would link patient-specific genetic variation to predictive models of protein function to inform treatment and guide therapy.
• Use computational models to predict clinical outcomes. An example may be the use of noninvasive imaging to obtain patient anatomy and computational fluid dynamics to predict hemodynamic and clinical outcomes of surgical reconstructive procedures.

Contributing Sources:

• NHLBI Strategic Plan Goals 1 and 2.
• NHLBI Strategic Plan Strategies 1, 2, 4, 6, and 8.
• Recommendations from the NHLBI Strategic Plan Level One Final Reports:
  • Bioinformatics and Computational Biology
• NHLBI Workshops and Working Groups
  • Modeling Mitochondrial Dysfunction in CV Disease Working Group, July 2007
  • Systems Medicine Working Group, January 2007

September 2008