Digital Drug Discovery

The Digital Drug Discovery program addresses a range of problems at the interface of chemistry, biology, modeling, data mining, engineering, and medicine, including medicinal chemistry and chemical biology. The program focuses on research questions relevant to the development of functional small molecules with specific physicochemical or biological properties; including protein-ligand interaction, property prediction, electronic structure modeling, chemical synthesis, materials, dynamics and kinetics, and basic drug discovery. The program also addresses chemical information analysis, data mining, and knowledge generation via semantic integration.

Vision

The Digital Drug Discovery program’s vision is to develop a translational drug informatics platform as a foundation to address the complex challenges in the development of chemical probes and human therapeutics, including accelerating the process and increasing the probability of success. Such a system must provide in-silico-analogous functionality of all aspects of an optimization cycle (testing, hypothesis development / refinement, synthesis, testing) in the different stages of (preclinical) development. It is built on various computational components, algorithms, data sources, ontologies, etc., which are integrated in a flexible modular architecture. The goal is to derive, capture, and effectively utilize knowledge from all accessible relevant internal and external data sources and tools, and from expert scientists.

Chemoinformatics Overview

Chemoinformatics and computer‐aided drug design methods play increasingly important roles in preclinical and translational drug research. To develop small molecule therapeutics and chemical probes, computational chemistry approaches are important to gain mechanistic insights and build models related to efficacy, ADME (absorption, distribution, metabolism, excretion), PK (pharmacokinetics), pharmacology, and toxicity. Chemoinformatics tools are also needed to analyze and extract knowledge from the enormous data sets that are generated by high‐throughput screening methods in the pharmaceutical industry, and also in the public domain (such as PubChem). The complexity of the drug discovery process manifests itself in high rates of clinical attrition primarily due to lack of efficacy and clinical safety (toxicity). Advances in systems biology and our increased knowledge in human genetics, functional genomics, and molecular biology hold the promise to expand the drug discovery paradigm from single‐target selective “blockbuster” drugs towards developing multi‐target drugs (polypharmacology) and individualized medicines.

The Frost Institute for Data Science and Computing’s goal is to integrate chemoinformatics, computational biology, and bioinformatics methods to develop a translational drug‐informatics platform as a necessary component to address these complex problems. The Drug Discovery program uses a distributed and parallelized computing environment for many of our modeling and data analysis procedures. On a number of projects, the team is working on innovative computational‐driven approaches and technologies that are relatively broadly targeted at the analysis and modeling at life science data with the goal towards developing small molecule chemical probes and human therapeutics.

For information on services and resources available through this program, please email idsc@miami.edu.

Stephan C. Schürer, PhD
Director, IDSC Digital Drug Discovery

Dr. Stephan C. Schürer joined UM in October 2008 to lead the Center for Computational Science’s (CCS) Cheminformatics program. His research then was centered in computer-aided drug design, cheminformatics, translational drug informatics, and semantic integration with the goal to better synergize experimental and ‘in-silico’ approaches for the development of small molecule tool compounds and drug “leads”.

More than a decade later, Dr. Schürer is an Associate Professor in the Department of Molecular and Cellular Pharmacology. The Schürer Lab focuses on Systems Biology Drug Discovery, Computational Chemistry, Medicinal Chemistry and Data Science to develop targeted translational therapeutics. The Medicinal Chemistry (MC) team utilizes tools and workflows designed to identify targets and hit compounds, to optimize these hits into advanced pre-clinical leads. MC is specifically interested in kinase drug discovery, where using machine learning approaches they’ve identified several first-in-class kinase molecular probes. This work is in collaboration with clinical and biomedical scientists at the Sylvester Comprehensive Cancer Center.

Dr. Schürer’s current research is focused on developing solutions large-scale integration and modeling of systems biology ‘omics’ and drug protein interaction data to guide translation of disease models into novel functional small molecules with a particular focus on kinases and epigenetic bromodomain reader proteins The Schürer research group is interested in developing novel approaches to integrate chemoinformatics, computational biology, and bioinformatics methods with medical chemistry to develop a translational drug-informatics platform to answer complex scientific questions.

Dr. Schürer is perhaps, best known for his work with the LINCS portal (Library of Integrated Network-Based Cellular Signatures), an NIH Common Fund program. LINCS has with the goal of generating a large-scale and comprehensive catalogue of perturbation-response signatures by utilizing a diverse collection of perturbations (e.g. chemical, genetic, disease state), model systems (e.g. cell lines, differentiated cells, embryonic stem cells) and assay types (e.g. gene expression, protein expression, epigenetic modification, imaging). Currently in Phase 2, LINCS consists of six Data and Signature Generation Centers (DSGCs) and one Data Coordination and Integration Center (DCIC) that together have produced over 400 datasets and over 50 analytical tools focusing on the deeper understanding of complex diseases and the development of novel and effective therapies. The cornerstone of this update has been the decision to reprocess all high-level LINCS datasets and make them accessible at the data-point level enabling users to directly access and download any subset of signatures across the entire library independent from the originating source, project, or assay.