This workshop focuses on the computational science aspects when navigating, exploring, or optimizing object functions in discrete Chemical Compound Space (CCS). Mathematical aspects of CCS deal with topics such as (i) isomerism and the development of topological descriptors or indices which find application in quantitative structure-property relationships, (ii) group theory, which finds applications in stereochemistry, quantum mechanics, and crystallography, and (iii) classification and characterization of subsets of CCS. Analysis and exploration of CCS requires computational science technologies like data-mining, organization analysis, optimization, and classification.
Applying these techniques requires the development of mathematical formulations that can be effectively analyzed with practical computational techniques. For example, the computational cost of evaluating properties of a compound can depend strongly on the combination of compositional variables that define the compound. Similarly, some CCS analysis problems can be approximated with continuous optimization formulations, which allows for fast analysis. Finally, parallel computing techniques will likely be needed to developing practical computational strategies for combining large-scale calculations of ensembles of compounds with iterative optimization algorithms.
This workshop will bring together experts in the fields of mathematical chemistry, biology, physics, materials sciences, and computational science, to report on recent research efforts that impact CCS analysis.
(Université d'Évry-Val d'Essonne)
William Hart (Sandia National Laboratories)
Peter Jones (Yale University)
Mauro Maggioni (Duke University)
Cynthia Phillips (Sandia National Laboratories)
Jean-Louis Reymond (Universität Bern)
Cenk Sahinalp (Simon Fraser University)
Mark Tuckerman (New York University)
Jean-Paul Watson (Sandia National Laboratories)