1 of 10

��REPORT ON ACTIVITIES OF THE �CROWD-SOURCING MODELS SUBGROUP��IMAG/MSM Working Group on �Multiscale Modeling and Viral Pandemics��December 2, 2021� �

Reinhard Laubenbacher

Laboratory for Systems Medicine

Department of Medicine

University of Florida

Gainesville, FL

reinhard.laubenbacher@medicine.ufl.edu

2 of 10

Group Focus:

The body’s response to a viral infection involves many different dynamic processes, at many different temporal and spatial scales. Individual laboratories typically focus on at most a few of these processes. In order to achieve effective models that can both advance our basic understanding of a viral infection at the systemic level and carry translational value it is important that individual models get integrated into a larger whole. This requires both a modeling and information infrastructure that makes this possible in an easy and distributed fashion, and effective collaboration, coordination, and communication between research labs. This subgroup of the MSM Working Group on Multiscale Modeling of Viral Pandemics will address all aspects of this problem.

Group leads: Laubenbacher, Macklin

3 of 10

Activities

Discussions around a collection of topics:

Obstacles to model sharing, reproducibility, adaptability
Obstacles to collaborating on model building and integration
What can we learn from existing community efforts?
Prepared an extensive set of notes

Eventually refocused on immune digital twins

Preparation of a manuscript

4 of 10

Building Digital Twins of the Human Immune System

G. An, J. Barhak, J. Glazier, T. Helikar, A. Knapp, R. Laubenbacher, P. Macklin, A. Niarakis, B. Shapiro, R. Sheriff, T.J. Sego

[final author order to be determined]

5 of 10

We propose that the coordinated development of high-resolution medical digital twins

can revolutionize the practice of both biomedical research and the delivery of medical care.

The aim of this article is to use the example of an Immune Digital Twin (IDT) to describe

key challenges for building and calibrating medical digital twins and how those challenges

can be met. We will describe both the high-level steps required for developing an IDT,

as well as highlight the need to develop a collaborative/crowdsourcing

environment/infrastructure necessary to bring together the multi-disciplinary

expertise required to achieve this goal.

6 of 10

Process of Developing an Immune Digital Twin (IDT):

Step 1. Construct a ”generic” template model of those parts of the

Immune system relevant for the particular application chosen.

Step 2. Personalize the template model to an individual patient.

How this is done will depend on whether the IDT is to be used online

or offline.

Step 3. Validation of the IDT

9 of 10

Action items

Organize an initial three-day workshop, bringing together representatives of all relevant

stakeholder communities, including systems immunologists, clinicians, computational

modelers (ranging from the molecular level to pharmacometrics), experts in information

management and curation, as well as informaticians. This group will develop specific

recommendations for each of the steps for immune system digital twin constructions

outlined earlier. This workshop will be repeated every year to update the roadmap,

discuss progress, and review new challenges and opportunities.

Once a conceptual map of a human immune system digital twin is established, work

closely with the systems modelling and systems biomedicine communities to identify

computational models that could be reused, infrastructure that could be reused and

adapted and identify clinical partners with access to patient cohorts and data to calibrate

and validate the component models.

Collaborate with consortia that work on similar projects building digital twins for other

types of health conditions (e.g., cardiac diseases, multiple sclerosis) to exchange lessons

learned and obtain feedback.

Identify possible funding sources, national and international, that could support the

entire project or individual investigators contributing to the immune system digital twin.

Adapt existing systems biology standards to address challenges of multimodal,

hybrid models and their accessibility by non-modelers.