Our research in biocomputing explores the use of Bacterial Gene Rgulatory Networks (GRNs) as intrinsic computational systems. Rather than engineering synthetic circuits, we harness the native GRNN architecture of bacteria such as E. coli to perform meaningful computation. Our key innovation lies in searching for application-specific sub-GRNNs within the natural network—enabling computation without genetic modification.

The inherent GRNN of E. coli can be viewed as a randomly structured neural network with approximately 5,000 nodes (genes) and over 10,000 regulatory edges, exhibiting diverse activation, inhibition, and combinatorial logic patterns. By identifying sub-networks that display computational potential, we extract regression- or classification-capable sub-GRNNs suited to specific tasks such as pattern recognition, signal transformation, or decision-making.