Systems Pharmacology Overview

Principles

Definition

Systems pharmacology is a research field and methodological approach that investigates the complex interactions between drugs and biological systems. It employs quantitative modeling, computational analysis, and the integration of multi-scale experimental data (from molecular to organismal levels) to understand drug mechanisms of action, predict therapeutic and adverse effects, and guide the discovery and development of novel therapeutics by considering the system-wide context of drug action.

Ontological type

Core Methods

Data Integration

Modeling Frameworks

Key developments

Key Figures

Computational Foundations era

Lucian L. Leape [1] affiliated with Harvard University [3] and Baylor College of Medicine [4] during the Computational Foundations era. The Systems Analysis of Adverse Drug Events [7] reframed safety as a process-level concern that linked molecular predictions to clinical and hospital-system outcomes. Ingo Muegge [2] collaborated with University of Southern California [5] and University at Buffalo, State University of New York [6] during this era. Simple Selection Criteria for Drug-like Chemical Matter [8] advanced drug-likeness filtering and enabled in silico lead discovery through practical descriptor-based screening.

Network Pharmacology Integration era

During 2003–2009, network pharmacology and systems pharmacology shaped predictive models by shifting emphasis from single targets to polypharmacology, with Andrew Hopkins advocating integrating network biology with descriptor-driven modeling to map multi-target actions. The network biology paradigm, as articulated by Albert-László Barabási and colleagues, provided the conceptual scaffolding for mapping drug actions onto interactomes, identifying network hubs and modules relevant to pharmacology. Descriptor-driven QSAR/QSPR and in silico ADME/safety profiling, advanced by George Ekins and collaborators, formalized preclinical risk assessment through defined applicability domains and external benchmarking to improve transferability across datasets. High-throughput screening and probabilistic classifiers for promiscuity and off-target effects, exemplified by Shoichet's work on promiscuous ligands and related approaches, operationalized systems-level assessments of drug behavior.

Multiscale Hybrid Systems era

Michael K. Gilson [1] is associated with Harvard University [3] and the Massachusetts Institute of Technology [4] in this era. BindingDB in 2015: A public database for medicinal chemistry, computational chemistry and systems pharmacology [7] provides a widely used resource for integrating data across drug discovery, pharmacology and translational modeling, crucial for multiscale predictions in this era. Yun Tang [2] is affiliated with East China University of Science and Technology [5] and Carnegie Mellon University [6] in this era. Her contributions include papers [8], [9], and [10], which collectively advance network-based, multiscale pharmacology and enable drug-target prediction, polypharmacology profiling, and GPCR-ligand modeling in this era.