PBET is a user-friendly computational tool designed to estimate the potential biological activity of phytocomplexes (mixtures of plant-derived compounds). It achieves this by:
Molecular Fingerprint Generation: PBET calculates Morgan fingerprints for each molecule within the phytocomplex. These fingerprints act as unique digital representations, capturing the molecules' structural features.Similarity Matching: PBET compares the generated fingerprints of the phytocomplex's constituent molecules against a library of known compounds with documented biological activities. The Tanimoto similarity coefficient, a well-established metric, is employed to quantify the degree of structural similarity between the phytocomplex molecules and the library entries.Data Integration and Analysis: PBET creates a comprehensive DataFrame that integrates the similarity scores obtained from the fingerprint comparisons. This DataFrame also incorporates additional valuable information, including target information, CAS numbers (unique identifiers for chemical substances), and the known bioactivities associated with similar compounds from the library.Visualization Potential (Optional): While not explicitly mentioned in the provided code, PBET likely offers functionalities for data visualization. This could involve generating charts or graphs to depict the relationships between the phytocomplex's components, their structural similarity to known bioactive compounds, and their predicted bioactivities.Accessibility:
The PBET notebook can be accessed and explored directly on Google Colab at the following link:
- https://colab.research.google.com/github/alessandrocareglio/PBET/blob/main/PBET.ipynb
For those interested in the underlying code and future modifications, the PBET codebase is also available on your GitHub repository:
- https://github.com/alessandrocareglio
Overall, PBET presents a valuable approach
