Natural product scores and fingerprints extracted from artificial neural networks

Due to their desirable properties, natural products are an important ligand class for medicinal chemists. However, due to their structural distinctiveness, traditional cheminformatic approaches, like ligand- based virtual screening, often perform worse for natural products. Based on our recent work, we evalu- ated the ability of neural networks to generate fingerprints more appropriate for use with natural prod- ucts. A manually curated dataset of natural products and synthetic decoys was used to train a multi-layer perceptron network and an autoencoder-like network. In-depth analysis showed that the extracted nat- ural product-specific neural fingerprint outperforms traditional as well as natural product-specific finger- prints on three datasets. Further, we explored how the activations from the output layer of a network can work as a novel natural product likeness score. Overall, two natural product-specific datasets were gen- erated, which are publicly available together with the code to create the fingerprints and the novel nat- ural product likeness score.