Methods and applications of data informed agent-based models for systems biology

Organizers:

Annequa Sundus (Indiana University Bloomington), Elmar Bucher (Indiana University Bloomington), Paul Macklin (Indiana University Bloomington)

Description:

Agent-based modeling is a powerful technique for spatial and temporal multiscale modeling of biological systems. It involves defining the agent, the rules the agents act on, and the physical and chemical environment information. Once the system is designed, then calibration is done on the emerging behavior of the system. Since agent-based models have inherent stochasticity along with a large number of parameters, model exploration needs a vast number of replicates for convergence.In practice, a coarse grain approach is often applied to very few parameters. Also agent-based tissue models are multiscale and can incorporate data from different sources from molecular dynamics to tissue scale imaging. However, the biggest challenge is to bridge the gap between data and simulation by using real-world data to inform and calibrate the models. Recent developments in spatial transcriptomics and image analysis have opened the possibility to better inform and calibrate agent-based models. In this mini-symposium we aim to present work that uses real-world data in the design and calibration of agent-based models. Our speakers will present methods and applications for using experimental data to inform agent-based models.

Diversity Statement:

We are proposing a mini-symposium featuring scientists at different career stages and from diverse institutions around the world. By bringing together a PhD student, a postdoc, and two faculty members, we aim to share unique perspectives on data calibration approaches from across all levels of academia. This mix fosters mentorship, collaboration, and a richer conversation, helping to build a more inclusive and innovative research community.