DataXflow: Synergizing data-driven modeling with best parameter fit and optimal control - An efficient data analysis...

Area of Science:

• Computational Biology
• Systems Biology
• Bioinformatics

Background:

• Data-driven models are crucial for extracting information from empirical data.
• Mathematical models encode relevant information by adapting parameters to specific data.
• Optimal control frameworks identify efficient targets for steering models via external stimuli.

Purpose of the Study:

• To introduce DataXflow, a software framework integrating data-driven modeling and optimal control for efficient target identification.
• To automate complex modeling tasks, including equation and script generation for systems with many agents.
• To lower programming barriers for researchers in systems biology and computational drug discovery.

Main Methods:

• DataXflow integrates three pipelines: D2D for model fitting, an optimal control framework for external stimuli, and JimenaE for graphical user interfaces.
• Equation generation from graphs and script generation for complex networks like gene regulatory networks.
• Iterative modeling process to refine model topology and generate regulatory networks from data.

Main Results:

• DataXflow successfully identified drug targets for lung cancer therapy, aiming to reduce proliferation and increase apoptosis.
• The optimal control framework revealed that inhibiting AURKA and activating CDH1 is the most efficient drug target combination.
• The framework facilitates an agile interplay between data generation and analysis for accelerated research.

Conclusions:

• DataXflow enables efficient drug target identification, even within complex biological networks.
• The software accelerates cancer research by providing a streamlined approach to therapeutic intervention modeling.
• This approach supports agile data analysis and potentially faster discovery of novel cancer therapies.