Phyto4Health

is a specialized database of secondary metabolites from medicinal plants included in the State Pharmacopoeia of the Russian Federation and is intended to support research aimed at discovering new natural-product-based pharmaceutical agents. The database contains information on 3,128 phytocomponents identified in 268 medicinal plants from the 14th edition of the Russian Pharmacopoeia. For each compound, the service provides:

Each plant is described in terms of taxonomy, species synonyms (533 synonyms were collected), and the phytochemical composition of individual plant parts, which distinguishes Phyto4Health from many analogous resources where compounds are listed for the whole plant without specifying the exact organ or tissue.

Biological activity prediction

For all 3,128 compounds, in silico prediction of biological activity was performed using PASS Refined 2022, which predicts 1,957 types of biological activity with an average accuracy of 97%. The PASS algorithm is based on a Naive Bayes classifier and MNA descriptors (Multilevel Neighborhoods of Atoms), and its output includes two estimates: Pa, the probability “to be active,” and Pi, the probability “to be inactive”. In total, the database contains more than 135,000 predicted records of pharmacological effects and mechanisms of action at the threshold Pa > 0.5.

Functionality

Database uniqueness

Comparison with three other natural product databases representing medicinal plants from different regions, SistematX (Brazil, 8,940 compounds), Ayurveda (India, 2,102 compounds), and NANPDB (North Africa, 6,328 compounds), showed that more than 57% of the structures in Phyto4Health are unique and differ substantially from phytocomponents found in those regional resources, with an average structural similarity of about 0.17 ± 0.09. This finding highlights the significant untapped potential of the Russian flora as a source of novel pharmaceutical agents. Thus, Phyto4Health is the first specialized tool for systematic in silico analysis of secondary metabolites from medicinal plants listed in the Russian Pharmacopoeia, integrating structural data, experimental activity data, and PASS predictions within a single user-friendly interface.

Practical Use

Phyto4Health can be used for virtual screening of plant-derived compounds, selection of candidate molecules with desired pharmacological profiles, and prioritization of compounds for experimental validation. The resource is also valuable for identifying phytocomponents linked to specific molecular targets, assessing possible mechanisms of action, and exploring the medicinal potential of individual plants or their specific parts. Because the database combines literature-derived phytochemical data with experimental target information and PASS predictions, it can support early-stage drug discovery, pharmacological profiling, and academic research on medicinal plants.