Drug Profile Matching (DPM) Method

In our Pattern-Based Drug Design platform, we develop a research service capable of discovering new effects of generic drugs and predicting side effects of drug candidates.

Most drugs exert their effects via multitarget interactions, as hypothesized by polypharmacology. While these multitarget interactions are responsible for the clinical effect profiles of drugs, current methods have failed to uncover the complex relationships between them. Our new approach is able to relate complex drug–protein interaction profiles with effect profiles. Structural data and registered effect profiles of all small-molecule drugs were collected. Interactions to a series of nontarget protein binding sites of each drug were calculated in molecular docking experiments, resulting the Interaction Profile (IP) Matrix.

Fig.1 Demonstration of the IP matrix. A drug molecule is docked to a set of 149 proteins, and the calculated binding free energies (docking scores) are entered into a row vector, i.e., the interaction profile (IP). IPs of the 1177 studied drugs form the IP matrix. Docking scores are shown in color code, blue denotes stronger interactions. On the right, 4 docked structures are presented.

Statistical analyses confirmed a close relationship between the studied 177 major effect categories and interaction profiles of ca. 1200 FDA-approved small-molecule drugs, and this relationship was confirmed by independent validation. Our results also show that the prediction power is independent of the composition of the protein set used for interaction profile generation. These findings allowed us to develop a robust prediction method named Drug Profile Matching, capable to reveal the effect profiles of drugs in their entirety, and to predict uncovered effects in a systematic manner.

Fig.2 Graphical summary of the Drug Profile Matching method: from the atomic structures to the effect probability matrix. The effect pattern (EP) matrix contains the therapeutic effects of the drugs in a binary coded form (blue and white cells represent the presence and the absence of a given effect from the 177 categories, respectively). Then, a canonical correlation analysis is performed with the IP matrix in order to generate highly correlating factor pairs that serve as the input for linear discriminant analysis. This way, classification functions are produced that yield the probability for each drug–effect pair, resulting in the effect probability matrix.

DPM is described in details in our paper “Drug Effect Prediction by Polypharmacology-Based Interaction Profiling” appeared in the Journal of Chemical Information and Modeling (JCIM) as a cover story. The issue is a sample issue of JCIM and the full text paper is available free of charge.

Research projects

ACTOMYOSIN ATOMIC STRUCTURAL MODELS

  • weak-binding actomyosin (actin trimer docked and relaxed with up lever Dictyostelium motor domain 1VOM), extra primed state
  • Download
  • activation loop mutant weak-binding actomyosin (actin trimer docked and relaxed with 1VOM R520Q mutant)
  • Download
  • loop 3 mutant weak-binding actomyosin (actin trimer docked and relaxed with 1VOM R562Q mutant)
  • Download
  • rigor actomyosin (actin trimer docked and relaxed with down lever Dictyostelium motor domain 1Q5G)
  • Download
  • rigor actomyosin (actin trimer docked and relaxed with down lever squid motor domain 2OVK)
  • Download