Investigation of the structural similarity of aromatase inhibitors developed for the treatment of breast cancer

Aromatase is a cytochrome P450 enzyme complex that serves a critical role in hormone regulation and various physiological processes by converting androgens into estrogens. In human breast cancer tissue, there is an overexpression of the aromatase enzyme. Aromatase inhibitors (AIs) are employed in the treatment of estrogen-dependent breast cancer in postmenopausal women to block the enzyme aromatase, thus inhibiting estrogen production. Undesirable effects associated with current AIs have necessitated an ongoing search for novel candidates with aromatase-inhibitory properties. This research focuses on the similarity of all known active AIs studied from 2010 to 2022, comprising 250 AI compounds collected from the literature through the Clarivate Analytics database (Web of Science). These compounds are described using various molecular descriptors and fingerprints, and then classified using K-means clustering and hierarchical clustering methods. Hierarchical clustering, in combination with ECFP fingerprints, produced promising results, dividing AI compounds into two main groups: steroidal and non-steroidal. The non-steroidal group is further subdivided into natural and synthetic compound groups. Using the results of ECFP fingerprints hierarchical clustering, we predict the antineoplastic activity of breast cancer via Passonline. Interestingly, we find that the group of natural compounds, all of which are active according to Passonline, is predominantly derived from Thailand. This methodology holds the potential to classify previously identified structures and serves as a valuable tool for evaluating active compounds on a given target.