Multiclass Molecular Knowledge Framework for Product and Process Design

Abstract

Computer Aided Product Design (CAPD) is widely used in process system engineering as a powerful tool for searching novel chemicals. The crucial steps in CAPD are the generation of candidate molecules and the estimation of properties, especially when complex molecular structures like flavors are sought. In this paper, we present a multiclass molecular knowledge framework which is based on chemical graph theory and chemical knowledge. Three kinds of functional groups are defined: elementary, basic and composed groups. These serve to generate four classes of knowledge that can be useful for property estimation and molecular design. An Input/output structure basing on XML language is defined to favour the interoperability between softwares.