NP-Scout: Machine Learning Approach for the Quantification and Visualization of the Natural Product-Likeness of Small Molecules

Abstract

Natural products (NPs) remain the most prolific resource for the development of smallmolecule drugs. Here we report a new machine learning approach that allows the identification of natural products with high accuracy. The method also generates similarity maps, which highlight atoms that contribute significantly to the classification of small molecules as a natural product or synthetic molecule. The method can hence be utilized to (i) identify natural products in large molecular libraries, (ii) quantify the natural product-likeness of small molecules, and (iii) visualize atoms in small molecules that are characteristic of natural products or synthetic molecules. The models are based on random forest classifiers trained on data sets consisting of more than 265,000 to 322,000 natural products and synthetic molecules. Two-dimensional molecular descriptors, MACCS keys and Morgan2 fingerprints were explored. On an independent test set the models reached areas under the receiver operating characteristic curve (AUC) of 0.997 and Matthews correlation coefficients (MCCs) of 0.954 and higher. The method was further tested on data from the Dictionary of Natural Products, ChEMBL and other resources. The best-performing models are accessible as a free web service at http://npscout.zbh.uni-hamburg.de/npscout.

NaPLeS: a natural products likeness scorer—web application and database

“…NP-scout [21] is, to our knowledge the only other web application allowing to compute an estimation on how much a submitted compound is likely to be a natural product. The approach used in NP-scout is different, as it computes a probability of a molecule to be a NP based on its physico-chemical properties, Morgan2 fingerprints and MACCS keys.…”

Cheminformatics in Natural Product‐based Drug Discovery

“…[16] NPs are on average heavier and more hydrophobic than synthetic drugs and synthetic, drug-like compounds. [59] Their structural complexity is also often higher, in particular with regard to stereochemistry (commonly quantified by the number of chiral centers, [57,[59][60][61][62][63][64][65][66] the number of fraction of Csp 3 atoms, [6,8] and/or the number of bridgehead atoms in ring systems [67] ) and 3D molecular shape. [8,68] NPs show an enormous diversity of ring systems, in particular of aliphatic systems.…”

Spectrum of deep learning algorithms in drug discovery

“…Chen et al developed a novel ML approach named NP-Scout with RF classifier for natural product-likeness quantification of small molecules. NP-Scout can also be used in the classification of small-molecule drugs for drug discovery (Chen, Stork, Hirte, & Kirchmair, 2019). Zhao and So employed various ML approaches to predict the drug repurposing probabilities for schizophrenia and depression disorders based on the expression profiles of available drugs.…”