SP 6
Mechanistic multiscale models
In this subproject a mechanistic multiscale model for aortic valve stenosis will be developed. This multiscale model integrates data and results from data mining on various levels (genome, hormone, cell, organ, patient, population) such that individual courses of disease and the efficacy of treatment concepts can be modelled. The objectives of this subproject are to provide individual predictions for patients participating in the SMART-sponsored clinical study of aortic valve stenosis.
The mechanistic multiscale model will include:
- A physiologically-based model of the cardio-vascular (CV) system
- A model of the renin-angiotensin-aldosterone-system (RAAS)
- A baroreceptor feedback-model
- A physiologically-based pharmacokinetic/pharmacodynamic (PBPK/PD) model of the concomitant pharmacologic therapy (beta-blocker, angiotensin receptor blocker, etc.)
- Correlations between measured lab parameters and pathophysiology that are discovered through statistical analyses.
This structure makes it possible to predict relevant cardiovascular parameters such as heart rate (HR) and cardiac output (CO)– even following surgical intervention and with concomitant pharmacologic therapy.
Optimization of therapeutic treatment will require an understanding of patient pathophysiology and its integration into the model. Therefore, a statistical analysis of correlations between “omics” data and pathophysiology will be performed. Data mining algorithms will be used for statistical analyses, and these will be adjusted and refined to the data within the framework of this project. Correlations that are identified will be incorporated into the structures of the individualized models to improve individual predictions.
Application of the established multiscale model will require parameterization and adjustment using data provided by the SMART project partners. To this end, „state-of-the-art“ algorithms in the field of Markov-Chain-Monte-Carlo (MCMC) methods will be used. This will enable a detailed, statistically rigorous analysis of predictive power, variability and uncertainty.
The multiscale model with the above-described features will be incorporated into the Computational Systems Biology Software Suite of Bayer Technology Services GmbH (www.systems-biology.com), which is designed specifically for applications in life sciences.
Keywords: PBPK, cardiovascular modeling, PK-Sim