Context-Aware Biology: GPCR Activation Prediction with Sequence Conservation Analysis

Current biological research workflows make use of disparate, poorly integrated systems that cause a large mental burden on the scientist, leading to mistakes on often long, complex, and costly experimental procedures. The lack of open tools to assist in the collection of distributed experimental conditions and data is largely responsible for making protocols difficult to debug and laboratory practice hard to learn. In this work, we describe an open Protocol Descriptor Language (PDL) and system to enable a context-rich, quantitative approach to biological research. We detail the development of a closed-loop pipetting technology and a wireless, sample temperature sensor that integrate with our Protocol Description platform, enabling novel, real-time experimental feedback to the researcher, thereby reducing mistakes and increasing overall scientific reproducibility.

We are working on prediction models for GPCR (G-protein coupled receptors) activation. Using olfactory receptors as a test case we are extracting information from primary sequence analysis, in order to pinpoint residues important for activation. Results are correlated with molecular dynamic simulations and in vitro experiments. GPCR activation is the major puzzle in drug discovery, where small molecules that activate only certain kinds of receptors are currently being found mostly by trial and error.