Journal Name:
The Open Nutraceuticals Journal
Volume:
3
Issue:
1
Pages From:
194
To:
202
Date:
Friday, January 1, 2010
Keywords:
Natural product, natural-likeness, drug-likeness, in silico prediction model, chemo-informatics
Abstract:
The majority of the currently used cosmetics and drugs are natural products-based compounds or their derivatives. This could add weight to the argument that natural based products are inherently better tolerated in the body than
synthetic chemicals and have higher chance to be approved as new drugs. The present study was undertaken to analyze a
natural product database compared to synthetic chemicals and to search for discriminative physicochemical properties that
may probably help in differentiating between natural and synthetic compounds. We have formulated rules to assess the
natural likeness of chemicals and thereby discriminate between natural-based and synthetic chemicals. A Mathews Correlation Coefficient of 0.5 was obtained; nearly 81% of natural-based products and 68% of synthetic chemicals were precisely classified using this filter. The property criteria for drug-likeness and lead-likeness are more pronounced in natural
products rather than synthetic ones. The fraction of synthetic chemicals which are natural-like could have higher chance to
be successful drug.