Collaboration and calibration: modelling with experimental and clinical data

Monday, June 14 at 5:45pm (PDT)
Tuesday, June 15 at 01:45am (BST)
Tuesday, June 15 09:45am (KST)

SMB2021 SMB2021 Follow Monday (Tuesday) during the "MS03" time block.
Note: this minisymposia has multiple sessions. The second session is MS04-IMMU (click here).

Share this


Adriana Zanca (The University of Melbourne, Australia), Jennifer Flegg (The University of Melbourne, Australia), Helen Byrne (University of Oxford, UK)


An ongoing challenge in mathematical and computational modelling is model validation and uncertainty quantification. Model calibration in the absence of rich clinical and experimental data relies on synthetic data, in turn adding different problems to solve. Conversely, clinical and experimental data can be so voluminous that it can be difficult to manage, interpret and use the data. Even where data allows for model calibration, uncertainty is often not analysed or reported. Without validation and uncertainty analysis, mathematical and computational models may not be clinically or experimentally applicable nor relevant. To help overcome this, there needs to be continual collaboration between clinicians, experimentalists, and modellers. This mini-symposium presents work across different fields and scales to highlight the range of experimental and clinical data available and how they can be best used in conjunction with mathematical modelling. Topics covered in this mini-symposium include angiogenesis, colon cancer, tumor modelling, ischemic heart conditions, lung disease, placental vasculature, HIV reactivation, and drug response.

Alison Betts

(Applied BioMath, USA)
"Modeling strategies for preclinical to clinical translation of T cell engager bispecific antibodies: using math to unravel counter intuitive dose responses"
T cell engager (TCE) bispecific antibodies are a promising therapeutic approach for the treatment of cancer. They have a complex mechanism of action, binding to CD3 on T cells and a tumor associated antigen on tumor cells to form a trimolecular complex (trimer), mimicking the normal immune synapse. Trimer formation stimulates the T cell and redirects cytotoxicity against the tumor cell. This results in some interesting mechanistic behaviors, including bell shaped concentration response relationships, which can result in non-intuitive dose response relationships. To understand these complex quantitative relationships, and to provide a tool for decision making from early discovery through to clinical trials, a translational quantitative systems pharmacology (QSP) model is proposed for TCE molecules.  The model predicts trimer formation between drug, T-cell and tumor cell, which can be linked to downstream pharmacodynamics, efficacy or toxicity. Two case studies are discussed; in the first the model is used to optimize design of a PSMA/CD3 TCE and in the second the model is used for preclinical to clinical translation of a Pcad/CD3 TCE to predict clinical efficacious dose.

Allison Lewis

(Lafayette College, USA)
"Bayesian information-theoretic calibration of tumor models for informing effective scanning protocols"
With new advancements in technology, we can now collect data describing tumor growth using numerous metrics. For any tumor growth model, we observe large variability among individual patients’ parameter values, particularly those relating to treatment response; thus, exploiting the use of these various metrics for model calibration can be helpful to infer such patient-specific parameters both accurately and early. Since clinicians are limited to a sparse collection schedule, the determination of optimal times and metrics for which to collect data in order to best inform model calibration is essential. Here, we employ a Bayesian information-theoretic calibration protocol for experimental design in order to identify the optimal times at which to collect data for informing treatment parameters. Data collection times are chosen sequentially to maximize the reduction in parameter uncertainty with each added measurement, ensuring that a budget of n measurements results in maximum information gain about the model parameter values.

Leili Shahriyari

(University of Massachusetts Amherst, USA)
"A data-driven mathematical model of colon cancer"
Every colon cancer has its own unique characteristics, and therefore may respond differently to identical treatments. Here, we introduce a data driven mathematical model for the interaction network of key components of immune microenvironment in colon cancer. We estimate the relative abundance of each immune cell from gene expression profiles of tumors, and group patients based on their immune patterns. We then compare the tumor sensitivity and progression in each of these groups of patients and observe differences in the patterns of tumor growth as well as response to FOLFIRI treatment.

Min Song

(University of Southern California, USA)
"Quantitative analysis of endothelial sprouting mediated by FGF- and VEGF-induced MAPK and PI3K/Akt pathways"
The essential role of blood vessels in delivering nutrients makes angiogenesis important in wound healing and tumor growth. Targeting angiogenesis is a prominent strategy in tissue engineering and cancer treatment. However, not all approaches to regulate angiogenesis lead to successful outcomes. There is a limited understanding of how pro-angiogenic factors such as VEGF and FGF combine together to stimulate angiogenesis. We aim to quantitatively characterize the crosstalk between VEGF- and FGF-mediated angiogenic signaling and endothelial sprouting, to gain mechanistic insights and identify novel therapeutic strategies. We constructed a hybrid agent-based model that characterizes endothelial sprouting driven by FGF and VEGF-mediated MAPK and PI3K/Akt signaling. The experimentally fitted and validated model predicts that FGF induces stronger angiogenic responses in the long-term compared to VEGF stimulation. Also, FGF plays a dominant role in the combination effects in endothelial sprouting. Moreover, the model suggests that ERK and Akt pathways and cellular responses contribute differently to the sprouting process. Furthermore, the model predicts that the strategies to modulate endothelial sprouting are context dependent. Thus, our model can identify potential effective pro- and anti-angiogenic targets under different conditions and study their efficacy. The model provides mechanistic insight into VEGF and FGF interactions in sprouting angiogenesis.

Hosted by SMB2021 Follow
Virtual conference of the Society for Mathematical Biology, 2021.