This presentation reports on the PKS Enumerator software to generate very large sets of virtual macrolides which are known to have valuable biological properties (<i>e.g., </i>erythromycin, solithromycin). As there is currently no such library in the public domain, our new cheminformatics software holds a unique potential in the field of computational chemistry and drug discovery for identifying new antibiotics and anti-cancer agents. The virtual screening of these complex molecules using computational models is highly relevant as the organic synthesis of such type of compound typically require at least 20 steps, thus dramatically slowing the discovery of new bioactive macrocycles. Our program allows users for a large variety of controls to design diverse or focused libraries of virtual macrolides (<i>e.g.,</i> types of extenders to be included in the macrocycle, global lengths of macrocycles, number of compounds generated, maximum number of stereocenters). To test and validate the software regarding the chemical diversity of generated macrocycles, we have enumerated three sets of 10 million diverse 12-extender macrolides using an initial set of 24 unique natural extenders. For each macrolide within each set, we calculated the molecular properties such as molecular weight, hydrogen bond donors/acceptors, number of rotatable bonds, or the topological polar surface area. The presentation discusses the chemical diversity of the libraries and their unique opportunities for virtual screening.

PKS enumerator to enumerate the chemical space of macrolides

254th National Meeting (2017)

The official archive and repository for all ACS Meetings & Expositions abstracts and select posters or presentations.