MSM in Drug Discovery and Development Theme

Return to 2014 Multiscale Modeling Consortium Meeting

MSM in Drug Discovery and Development (David D’Argenio lead)

Multiscale modeling may provide the integrating framework needed to understand the therapeutic and toxic effects of drugs over multiple spatial and temporal scales, and contribute to a unifying framework to bridge the discovery/development and translational barriers currently limiting the development of new medicines. Two NIH sponsored workshops held in 2008 and 2010 and organized by Michael Rogers (NIGMS), along with Paul Brazhnik (NIGMS), Jennifer Couch (NCI), Sarah Dunsmore (NIGMS), Peter Lyster (NIGMS), Dick Okita (NIGMS), and Grace Peng (NIBIB), led to the white paper entitled, Quantitative and Systems Pharmacology in the Post-genomic Era: New Approaches to Discovering Drugs and Understanding Therapeutic Mechanisms, Peter Sorger et al., Oct. 2011 ( This document begins to codify the role of quantitative and systems pharmacology as a framework for discovery, development and rationale use of therapeutics, by presenting directions for future research as well as for education and training.

Since these two NIH workshops and the publication of the roadmap outlined in the white paper by Sorger et al., there has been a range of activities under the rubric of Systems Pharmacology, including: workshops (e.g., Am. Soc. of Clinical Pharmacology Workshop on Quantitative and Systems Pharmacology, March 2013), a new journal, (e.g., CPT: Pharmacometrics & Systems Pharmacology, inaugural issue September 2012); monographs (e,g., Systems Pharmacology and Pharmacodynamics by D. Mager and H Kimko, which will include a section on  multi-scale models of drug action, publication 2014), funding initiatives (e.g., RFA:AG14-017 Planning Grants for Alzheimer's Disease Translational Centers for Predictive Drug Development), and numerous others.

The overall goal of the 2014 MSM Consortium theme on Multiscale Modeling in Drug Discovery and Development is to elaborate the role of multiscale modeling as the platform for vertical integration that seeks to synthesize information of drug action at the molecular, sub-cellular, cellular, multi-cell, tissue, organ, organism, and population levels.

Schedule--Thursday, September 4th

9:00 David D’Argenio, Univ. of Southern California, Systems Pharmacology as an Integrating Framework for Drug Discovery and Development

9:00-9:20 Colleen Clancy, University of California, Davis, Predicting the Emergent Effects of Antiarrhythmic Drugs on Cardiac Rhythms

9:20-9:40 Stacey Finley, Univ. of Southern California, Anti-angiogenic Cancer Therapies Targeting the VEGF Pathway

9:40-10:00 Susana Neves, Mount Sinai School of Medicine, Phophodiesterase Control of AMPAR Trafficking

Abstracts and Bios

David D'Argenio

David D'Argenio is Professor of Biomedical Engineering at the University of Southern and holder of the Chonette Chair of Biomedical Technology. He is a Fellow of the American Institute for Engineering in Medicine and Biology, the American Association of Pharmaceutical Sciences, and a past member of the FDA Advisory Committee for Pharmaceutical Science and Clinical Pharmacology. His research focuses on the development of modeling methodologies for PK/PD systems analysis including: experiment design for modeling sparse data systems; computational Bayesian methods in PK/PD; adaptive control of drug therapy; population modeling methods; construction of genetic networks from heterogeneous data sources; and computational platforms for PK/PD systems modeling. These methods have been applied to numerous problems in drug development in the areas of oncology, infectious diseases, metabolic disorders and immunology.

Colleen Clancy - Predicting the Emergent Effects of Antiarrhythmic Drugs on Cardiac Rhythms

Common electrical diseases like cardiac arrhythmia that affect millions of Americans are notoriously difficult to manage with drug therapy, and some drugs intended for therapy can even exacerbate disease. A vital hindrance to safe and effective drug treatment rhythm disorders is that there is currently no way to predict how drugs with complex interactions and multiple subcellular targets will alter the emergent electrical activity of cells and tissues. We are developing a computer based system predictive multiscale model framework that will allow for probing the mechanisms of action of drugs and predicting their emergent effects in the heart.


Colleen E. Clancy received her Ph.D. in Physiology and Biophysics from Case Western Reserve University and is currently a Professor of Pharmacology at the University of California Davis. She serves on the advisory board of the National Biomedical Computation Resource. Her key areas of research include 1) modeling of nonlinear excitable systems; (2) computational pharmacology; (3) modeling and simulation of reaction and diffusion; (4) computational approaches to link genotype to phenotype.

Stacey Finley - Anti-angiogenic Cancer Therapies Targeting the VEGF Pathway

Vascular endothelial growth factor (VEGF) is a potent regulator of angiogenesis whose role in cancer biology has been widely studied. VEGF has been targeted in various cancer treatments, and anti-VEGF therapy has been used clinically for treatment of several types of cancer. Systems biology approaches, including computational models, provide a framework to test biological hypotheses and optimize effective therapies that aim to inhibit tumor vascularization and growth. In this work, we investigate the effects of anti-angiogenic therapies targeting VEGF and its receptors using a molecular-detailed compartment model of VEGF kinetics and transport in the human body. The model includes two major VEGF isoforms (VEGF121 and VEGF165), receptors (VEGFR1, VEGFR2), coreceptors (Neuropilin-1 and Neuropilin-2), and soluble factors (sVEGFR1 and -2-macroglobulin). We apply the model to simulate the effects of therapeutics that target VEGF. Interestingly, the model predicts that VEGF in the tumor interstitium can increase or decrease following administration of the VEGF antibody, depending on properties of the tumor microenvironment. The model is useful for understanding the dynamics of VEGF distribution in the body in response to anti-VEGF agents. Thus, our model generates clinically relevant predictions in the areas of drug mechanism of action, biomarker identification, and personalized medicine.


Stacey Finley received her Bachelor's degree in Chemical Engineering from Florida A & M University in 2004 and her Ph.D. in 2009 from Northwestern University. Following her graduate work, Dr. Finley was a postdoctoral research fellow at the Johns Hopkins University School of Medicine. She was awarded postdoctoral fellowships from the NIH National Research Service Award and the UNCF/Merck Science Initiative. Dr. Finley joined the Biomedical Engineering Department at USC in the summer of 2013. Dr. Finley's research program applies a systems biology approach to develop molecular-detailed computational models of biological processes related to human disease, and she is particularly interested in angiogenesis and cellular metabolism, two processes that are essential to cancer growth and other diseases.

Susana R. Neves - Phophodiesterase Control of AMPAR Trafficking

Neurons regulate their synaptic communication (synaptic strength) by changing the number of postsynaptic neurotransmitter receptors. cAMP and cGMP, the ubiquitous second messengers of the brain, transduce signals of neuromodulators into synchronized regulation of neuronal excitability by regulating receptor trafficking and changing synaptic strength. The modulation of synaptic strength is dysregulated in a number of psychiatric and neurological disorders, thus identifying targets that reverse these maladaptive changes could have therapeutic value. Phosphodiesterases (PDEs) are essential in shaping the spatial and temporal dynamics of cAMP and cGMP signaling, making them promising therapeutic targets for cyclic nucleotide driven neuropathologies. We find that the complex cross-regulation of multiple PDE activities can produce unexpected and at times counterintuitive signaling outcomes. We have developed dynamical computational models of signaling to understand this interplay of PDE activities, as we believe it could account for some of the failures of individual PDE inhibitors in clinical settings. We hypothesize this interplay of PDE activities could be exploited to eventually develop a poly-pharmacological approach to repurpose failed PDE inhibitors to treat a number of psychiatric and neurological diseases.

Susana R. Neves is an assistant professor in the department of Pharmacology and Systems Therapeutics at the Icahn School of Medicine at Mount Sinai. Her research interest is in understanding the contribution of phosphodiesterases to synaptic responsiveness using a systems biology approach combining dynamical computational models of signaling with imaging-based methods that measure cyclic nucleotide signaling and receptor trafficking in live neurons.

Breakout Session (Thurs., Sept. 4, 12:30-1:30pm)

Session Facilitators: Aleksander Popel and David D'Argenio

Quantitative and Systems Pharmacology in the Post-genomic Era: New Approaches to Discovering Drugs and Understanding Therapeutic Mechanisms, Peter Sorger et al., Oct. 2011, listed the following goals for a systems pharmacology. How can multiscale modeling contribute to achieve\ing these objectives?

  • Characterizing the biochemistry of drug targets, their networks and the effects of drugs
  • Investigating the origins of variability in drug response in single-cells, organs and patients at the level of the proteome, genome and environment
  • Exploiting diverse clinical and omic data to create pharmacodynamic biomarkers that inform integrated, multi-scale models of drug response in distinct patient populations
  • Developing better animal/tissue models for improved target validation
  • Reconnecting tissue physiology with chemistry to facilitate pharmacological experimentation and phenotypic screening on cells and model organisms
  • Developing and supporting information exchanges for QSP, particularly in the area of clinical data and electronic medical records
  • Developing multi-scale computational models of pharmacological mechanism that can bridge cell-level biochemical and organism-level PK/PD processes•Developing  formalized approaches for the analysis of failed clinical trials


Other areas for possible contribution of MSM:

       In drug discovery:

  1. Develop. of molecular-detailed pharmacokinetic models
  2. Targeting multiple pathways with a single or multiple agents (monotherapy vs combination therapy)
  3. Predicting drug toxicity
  4. Biomarker discovery

       In drug development:

  1. Role of MSM in drug formulation development, incl. nano
  2. Using MSM in design and analysis of clinical trials
  3. Patients population stratification


Notes for MSM in Drug Discovery and Development Breakout Session

The following is compilation of notes kindly prepare by both Stacey Finley and Feilim Mac Gabhann. The comments recorded in these notes are presented below without attribution.

  • Vertical and horizontal integration are essential components of the systems pharmacology for drug discovery and development. Multiscale modeling (across scales, targets/cells) is an essential component in this effort
  • Goal of the breakout is to nucleate a working group in this area
    • Open to ideas for name, goals, structure
  • Drug development implies a target is already identified
    • How can multiscale modeling contribute to target identification?
    • How can we formalize this?
  • Reason to study known compounds, especially failure: drug repurposing; better stratify for new compounds coming into the pipeline; rehabilitate the drug to optimize pros and minimize cons; failure analysis, with access to the clinical data of the failed trial; connect computational arm to patient-specific data
    • Need failure analysis but grants keep getting rejected - 'we know they don't work'
    • Need data from trials to do failure analysis
  • What is missing: chemistry behind the target –which features give it what characteristics; chemical structure is a data set – what does it tell us
  • National Centers for Systems Biology have little reference to chemistry, and they should. It's a piece missing from the development of systems biology approaches
  • Omics identification of targets alone is not effective
    • Need more understanding beyond up- and down- regulation
    • Better analysis, with mechanistic detail, within the context of the known biology
      • Can this methodology be codified?
      • Need to know the interface between omics and mechanisms
      • Difficult to validate: omics -> outcome
    • Sorger provided a proof of concept combining mechanism with disease function
    • Need to connect Omics to hard-won biomechanism information, instead of looking only at linear models of omics alone
    • Enhanced pharmacodynamics (Iyengar and colleagues)  offers one approach
  • Development of centralized repository of models
    • No reason to not have a standardized PBPK model
    • compilation of the models
    • Along with recorded seminar to describe the model (JOVE, workflow)
    • Shared goal with other multiscale working groups, with all of the same challenges.
  • Is there a role of MSM in drug formulation? Design of clinical trials? Stratify patient populations?
    • Yes! Look at multiple scales and processes
    • Since the models would be used in this way, should the models become FDA-approved?
      • Most of the responsibility falls on the sponsor – what does the FDA need to make the decision
      • Best practices for analyzing use of models
  • Biomarkers
    • MSM can predict which combination of multiple factors are predictive
      • Example: DREAM challenge, predict fast- and slow- ALS progression
  • Incorporating disease progression into the models
    • Incorporate disease heterogeneity
      • Different cell types and interactions – probably need a 3D model
      • Could link to a compartment model for metastases, circulation, drug PK
      • Pick a disease or pick a common problem across diseases (liver toxicity)
  • Focus on liver toxicity / profile: common to all diseases (rather than disease-specific models
    • Need to reach out to pharmaceutical industry
    • FDA has QSAR models to predict toxicity – can extend to metabolites that limit drug efficacy
  • Need to reach out to pharmaceutical industry
    • Why isn’t there rep from computational chemistry?
      • Need to look at atomic level
  • Need to include the informed clinician and patient advocacy perspective
    • Present challenges with treatment. This will help disrupt the linear one-way paradigm
    • Becuase of its broad undestanding of systems at all levels, the MSM community is uniquely capable of important contributions to improve how new medicines are made available and used to treat diseases


Pharmakinetic modeling software

Submitted by yalingliu on Thu, 09/04/2014 - 11:27

Table sorting checkbox