Press Release: Combinatorial Synthetic Biology and a Revolution in AI Agent Biomodelling at SynBioBeta ‘26

Dr. Eric W. Thomson-Wasiolek will be introducing and speaking on Combinatorial Synthetic Biology and

an AI Agent Modeling in general with a specific example of Intercellular Signaling modelling and

simulation at SynBioBeta’s Lunch and Learn at 12:10 on Tuesday May 5, 2026.

Combinatorial Synthetic Biology uses AI and computers to search through massive possibility spaces to

identify novel proteins with biologically active binding pockets, novel genes which encode novel

proteins, and novel synthetic pathways. It is estimated that there are 1082 atoms in the universe.

However this number is minute compared to the possible number of synthetic proteins of average

length of 350 amino acids that can be produced (20350 = 2.29 * 10455) and the possible number of

synthetic genes that can be produced (41050 = 1.46 * 10632) of average coding region (1050 nucleotides)

and possible number of synthetic pathways that can be generated (approximately 1062) with the

average pathway being 15 enzymes out of 15,000 average number of enzymes in a cell. Combinatorial

synthetic biology possibility spaces can be precisely quantified.

These possibility spaces are huge and can be radically reduced by imposing successive constraints on the

possibility spaces to make them computable, such as identifying only synthetic proteins that have

biologically active binding pockets (1 in a billion) or have specific binding pocket features. Computing

possible synthetic proteins, genes, and pathways enables a rapid expansion of possible synthetic biology

solutions to various corporeal conditions and diseases.

The second part of this presentation will introduce agentic AI architectures for biology. A particular

example of an application generated by the presenter on inter and intracellular signaling will be

presented. This architecture involves 10 agents which simulate the sending of cell signals to receptors

that initiate intracellular pathways which generate a cell fate (either apoptosis, gene expression

differentiation, or cell proliferation). The simulation is dynamic, with multiple signals sent to multiple

receptors, initiating multiple signaling pathways asynchronously in a dynamic simulation. Moreover this

application simulates six types of cell signaling: autocrine, where a cell sends a signal to itself, paracrine

where a cell sends signals to other cells in a diffusion area, juxtacrine signaling where a cell's signal is in

contact with the receptor of another cell, endocrine where a signal is sent via the blood stream to a

distant population of cells, gap junction signaling where ions are sent from one cell to another, and

neurotransmission where a neurotransmitter is sent from the pre-synaptic site to a post-synaptic site.

Specific examples of signaling simulations will be presented involving all three cell fates, medicines to

induce apoptosis or cell cycle arrest, and a synthetic biology simulation will be presented where a

synthetic pathway produces a certain molecular target.

The intercellular agent architecture here uses bio-causal reasoning, reinforcement learning, and

differential equation modeling of molecular dynamics.The speaker has a doctorate in computer science, a masters in computational molecular biology with a

masters thesis done at Stanford in a computational neuroscience project where he wrote seventeen C++

programs using a graph data structure to model the neurogenetic and synaptic development of the C.

elegans brain, as well as an MBA, and degree in Philosophy from U.C. Berkeley. Dr. Thomson-Wasiolek

is the owner and founder of Biomedical Reasoning Systems.