Chemical space and diversity of seaweed metabolite database (SWMD): A cheminformatics study

Publication: J Mol Graph Model
Software: ADMET Predictor®

Abstract

Seaweeds have attracted attention in the past decade as a biological source of highly diverse secondary metabolites with great potential in the industrial and pharmaceutical sciences. Herein, we represent a comprehensive cheminformatics study to compare the chemical diversity of seaweed metabolites based on their taxonomic source. Seaweed Metabolite Database (SWMD) was utilized in this study. The compounds were manually categorized into three datasets, namely red algae (Rhodophyta, n = 645), brown algae (Phaeophyta, n = 220), and green algae (Chlorophyta, n = 32). The compounds in each dataset were curated to generate six chemical descriptors of pharmaceutical interest for each molecule, which were later used to visualize the chemical space of these metabolites by principal component analysis. Scaffolds were generated by removing side chains and keeping the core part of each molecule. Scaffold diversity among the tested datasets was quantified using Cyclic System Retrieval Curves. Green algae metabolites in SWMD possessed the highest scaffold diversity followed by brown and red algae metabolites, respectively. Three structural binary fingerprints, including ECFP_4, MACCS keys, and PubChem were computed indicating that the red algae metabolites had the highest fingerprint diversity followed by the green and brown algae metabolites respectively. Finally, Consensus Diversity Plots were generated to assess the global diversity considering both scaffold and fingerprint diversity. It was concluded that green algae metabolites in the SWMD are the most diverse regarding chemical descriptors of pharmaceutical relevance and scaffolds. While red algae possess the highest fingerprint diversity.

By Ahmed H. Al Shariea, Tamam El-Elimat, Yazan O.Al Zu’bi, Abdelwahab J. Aleshawi, José L. Medina-Franco