molecular computing

Thursday Extra on 9/14

Thursday, September 14, 2017
4:15 p.m. in Science 3821
Refreshments at 4:00 p.m. in the Computer Science Commons (Science 3817).
Everyone is welcome to attend!

Cauldron: An IDE for modular development of chemical reaction networks will be presented by Prof. Klinge's summer research group: Jong Hoon Bae, Theo Kalfas, Nick Roberson, and Otabek Nazarov.

The chemical reaction network (CRN) model is a prominent molecular programming paradigm. In 2016, Klinge, Lathrop, and Lutz introduced three ways to modularly develop CRNs: I/O CRNs, extension operators, and closed sub-CRNs. I/O CRNs extend the CRN model to allow time-varying input concentrations. Extension operators are used to automatically add functionality to a CRN without affecting the original behavior. A closed sub-CRN encapsulates a behavior within an existing CRN that is self-contained. These new methods naturally support modular CRN design; however, existing tools do not support them. In this talk we introduce Cauldron, an integrated development environment (IDE) for modular CRN development. Cauldron supports I/O CRNs, extension operators, and closed sub-CRNs in addition to common features found in existing CRN tools. For example, users can divide a CRN into independent sub-CRNs, test them separately, and reuse them in other CRNs. Furthermore, users can mark species as inputs and specify them with common elementary functions, by drawing a function, or by connecting them to another CRN. Many commonly used CRNs and extension operators are also included as libraries in Cauldron. (This is joint work with James Lathrop at Iowa State University.)

Thursday Extra 9/22: Molecular Computing

Thursday, September 22, 2016
4:15 p.m. in Science 3821
Refreshments at 4:00 p.m. in the Computer Science Commons (Science 3817)

Preventing Memory Corruption in Chemical Computations
Professor Titus Klinge will discuss his recent work on molecular computing. Molecular computing systems that are contained in well-mixed volumes are often modeled using chemical reaction networks. In these systems, concentrations of molecules are treated as signals and used for both communication and memory storage. A common design challenge for such a system is to avoid memory corruption caused by noise in the input signals. In this talk, he overviews recent results concerning two related signal restoration algorithms for molecular systems modeled with chemical reaction networks. These algorithms are designed to prevent a memory signal from degrading over time, and he shows that under modest conditions these algorithms will maintain the memory indefinitely.

Syndicate content