The Concept and Scientific Vision of Way2Drug

The Way2Drug platform is founded on the integration of structure-property and structure-activity relationship methodologies with advanced machine learning techniques, enabling the systematic prediction of biological activity profiles, as well as key pharmacodynamic and pharmacokinetic parameters of chemical compounds, based solely on their structural formula. The underlying premise is that the accumulated experimental data on hundreds of thousands of previously studied substances can be transformed into mathematical models ((Q)SAR/SSPR), which then enable the evaluation of new molecules' properties prior to their synthesis and biological testing.

In this way, the platform serves as a bridge between cheminformatics and experimental pharmacology: it helps translate structural data about a molecule into predictions of its pharmacological effects, mechanisms of action, toxicity, metabolism, and potential therapeutic indications. This concept reflects the broader shift in modern drug development from purely empirical discovery toward rational, data-driven drug design.

Structure and Key Services of Way2Drug Platform

Way2Drug is a suite of interconnected web applications developed by an interdisciplinary team of specialists in bioinformatics, cheminformatics, and computer-aided drug design. The platform offers services for predicting thousands of types of biological activity, toxic and side effects, interactions with enzymes and transporters, effects on gene expression, and other parameters critically important for evaluating the profile of a drug candidate.

The central service is the PASS Online web service (Prediction of Activity Spectra for Substances), which predicts more than 4,000 types of biological activity from a structural formula, including pharmacological effects, mechanisms of action, toxic and adverse effects, interactions with metabolic enzymes and transport proteins, as well as effects on gene expression. To use PASS Online, it is sufficient to specify the structure of a compound (including one that has not yet been synthesized), after which the service will generate a spectrum of probabilities for different types of activity with high average prediction accuracy, confirmed by cross-validation.

Our platform includes a number of specialized services focused on individual aspects of pharmaceutical development. These include:

Of particular interest is a web service MetaPASS for analyzing the spectrum of biological activity with consideration of the metabolic transformation of compounds. It uses data on the metabolic pathways of thousands of substances from ChEMBL, DrugBank, and other databases, making it possible to assess not only the activity of the parent compound but also the contribution of its metabolites to the overall pharmacological and toxicological profile. Additional platform resources provide capabilities for storing and retrieving structures, working with compound sets, and facilitating interaction among members of the Way2Drug scientific community. You can learn more about all our services in the relevant section.

Opportunities for Scientists and Practical Applications

For scientists engaged in drug discovery and development, Way2Drug is a multifunctional web platform supporting various stages of R↦D from virtual screening and hit identification to repositioning of already approved drugs. The capability to simultaneously predict thousands of biological activities across structurally diverse compounds enables a comprehensive assessment of their pharmacological potential and supports the rational prioritization of the most promising candidates for subsequent experimental validation.

The platform is useful both when working with new virtual structures and when re-evaluating existing compounds, including approved drugs and natural products. Predicting activity spectra helps identify unexpected potential effects and new mechanisms of action, creating opportunities for drug repositioning, or finding new therapeutic uses for registered drugs.

In addition, early computational identification of likely side effects, toxicity, and undesirable target interactions helps reduce the risk of late-stage failure of candidates during preclinical and clinical studies.

An example of practical use of the platform is virtual hit screening from large compound libraries, where PASS Online and other services are used to compile a list of molecules with the most suitable combination of probable pharmacological effects and minimal risk of adverse properties. In another typical scenario, a researcher uses Way2Drug to analyze an existing compound collection in order to discover previously overlooked activities and, on this basis, propose new directions of application or drug combinations.

The Role of Way2Drug in Modern Pharmaceutical Development and the Scientific Community

In an environment where the cost of developing a new drug is measured in billions of dollars and the probability of success across all stages of development remains low, the use of computational platforms such as Way2Drug becomes an important tool for reducing costs and risks. By providing free (for academic users) access to predictive models, the platform contributes to more rational planning of synthesis and biological experiments, focusing resources on compounds with the greatest chance of success.

The mission of the Way2Drug team is defined as creating a favorable environment for effective collaboration among specialists working in the interdisciplinary field of drug development. The portal not only provides computational services but also serves as a platform for data sharing, accumulation of structures and predictions, and integration of various computer-aided drug design methods. As a result, Way2Drug becomes not merely a set of programs, but an element of the infrastructure of the modern scientific community, uniting medicinal chemists, pharmacologists, bioinformaticians, and machine learning specialists.

Looking ahead, the development of such platforms leads to closer integration of experimental and computational approaches, where the results of virtual screening are promptly verified in the laboratory, and new experimental data are used to improve models and expand predictive capabilities. Way2Drug, fulfilling the role of an informational and computational hub, contributes to the formation of this closed-loop cycle of "data - model - prediction - experiment," which ultimately improves the efficiency and predictability of new drug development.