FAST Congress  

2014 Phenoypic Drug Discovery
Download Brochure | Register 

Phenotypic screening (aka classical pharmacology) has been historically used in drug discovery. While technological developments have made the prevalence of target-based screening more popular, statistical analysis shows that a disproportionate number of first-in-class drugs with novel mechanisms of action come from phenotypic screening. Cambridge Healthtech Institute’s Second Annual Phenotypic Drug Discovery meeting will address the advantages of phenotypic screening vs. target-based screening, and focus on assay development, selection of physiologically-relevant models and subsequent target identification.

Sunday, November 16

5:00 pm Short Course Registration and Main Conference Pre-Registration

Monday, November 17

7:00 am Conference Registration and Morning Coffee

Case Studies in Phenotypic Drug Discovery

8:00 Chairperson’s Opening Remarks

Regis Doyonnas, Ph.D., Pfizer

8:10 Phenotypic Drug Discovery Advances toward Pharmacological Chaperone Therapies

Regis Doyonnas, Ph.D., Senior Principal Scientist, High-Content Screening and HTS-Flow Cytometry, Hit Discovery and Lead Profiling, Worldwide Research & Development, Pfizer

Pharmacological chaperones correct the proper folding of misfolded proteins protecting them from degradation and allowing them to correctly localize within the cell. For many years, drug discovery assays have focused on direct binding of pharmacological chaperones to their protein targets. The relative success of this approach has been called to question with the increasing discussion and analysis of the merits of phenotypic screening. While pharmacological chaperone assays may not lend themselves as “canonical” phenotypic screens, i.e., an assay for direct disease associated outcomes, a case can be made that broadening the analysis to a biological read-out resulting from chaperone activities within a cellular environment would reflect more of the pertinent biology of the system and thus increase our level of confidence in translation to patients.


8:35 Phenotypic Screening: A New Perspective

Jing LiJing Li, Ph.D., Director, Genomics and Phenotypic Screening, Merck Research Laboratories

Two general approaches of drug discovery, either target-centric or phenotypic, are usually taken. Interestingly, in the first-in-class drug category, the phenotypic approach yielded more approved drugs than the target-centric approach during the period 1999-2008. The lack of chemistry support and the immaturity of technology platforms for protein target identification have contributed to the low success rate for past phenotypic screens. Recent advances in the fields of affinity capture, quantitative mass spectrometry, and chemoinformatics greatly improve our chances to identify the underlying protein targets. This provides us opportunities to revisit this subject. With the lessons learned, the possibility of successfully applying phenotypic screens in drug discovery can improve significantly.

9:00 Phenotypic Discovery at AstraZeneca

AnnaCollenAnna Collén, Ph.D., Director, Reagents and Assay Development, AstraZeneca R&D, Mölndal, Sweden

A key denominator for successful phenotypic screening is the translational link of the in vitro cell systems and assays to eventually mimic human physiology. This will direct the positive outcome of the phenotypic screen and enable target identification of relevance to drug discovery. We will present how we at AstraZeneca have worked with phenotypic discovery in different areas: neuroscience with ApoE phenotypic screen, regenerative medicine, cardiac regeneration, islet health and finally identification with factors that differentiate white adipocytes to “brownish.”

9:25 Identification and Optimization of a Novel Phenotype Screen Hit to the First-in-Class HCV NS5A Inhibitor Daclatasvir

Makonen Belema, Ph.D., Senior Principal Scientist, Virology Chemistry, Bristol-Myers Squibb

High-throughput screening of the BMS compound collection afforded a mechanistically unique thiazolidinone hit exhibiting a sub-micromolar inhibitory potency towards an HCV replicon. Raising and mapping of resistance mutation indicated that the NS5A protein, a key protein with a multifunctional role in the virus’ replication cycle, was the most likely target. Highlights of the medicinal chemistry campaign that optimized this screen hit to the highly potent first-in-class HCV NS5A inhibitor daclatasvir along with key clinical results will be discussed. In addition, aspects of the mechanistic study that not only corroborated target engagement but also resulted in the discovery of a class of NS5A-targeting molecules that synergize the inhibitory potency of daclatasvir toward resistant mutants will be covered.

Retrogenix9:50 Deconvoluting Receptors Targeted by Phenotypic Screen Hits Using the Retrogenix Cell Microarray Platform
Britz_MattMatt Britz, Director, North American Business Development, Retrogenix
Selecting small molecule hits by phenotype, rather than against a predetermined target, leads to novel, disease-relevant targets. Retrogenix’s Cell Microarray technology overcomes the difficulties of target deconvolution with high success rates. It also provides a powerful approach to identify potential off-target activities to guide lead selection, and identifies previously unknown binding partners. 

10:05 Coffee Break in the Exhibit Hall with Poster Viewing 

10:35 Factors Important to a Successful Phenotypic Screening Strategy

David C. Swinney, Ph.D., CEO, Institute for Rare and Neglected Diseases Drug Discovery (iRND3)

Phenotypic assays are essential tools for drug discovery, with different endpoints depending on the goals: 1) empirical endpoints to understand the underlying biology, 2) empirical endpoints to identify toxicity of drug candidates, and 3) knowledge-based endpoints (biomarkers) to guide discovery which are ideally translational. The value of phenotypic assays is increased through effective alignment of phenotypic assay endpoints with the objectives of the relevant R&D stage.

11:00 Heterogeneity in Drug Discovery, Development and Diagnostics

LansingTaylorD. Lansing Taylor, Ph.D., Director, University of Pittsburgh Drug Discovery Institute and Allegheny Foundation Professor, Computational and Systems Biology, University of Pittsburgh

Heterogeneity is a fundamental characteristic in biological systems. Heterogeneity can arise from genetic variations, as well as from non-genetic factors. The ability to detect and to quantify heterogeneity is critical for successful drug discovery, development and diagnostics, and high-content analyses (HCA) permit the detection and quantitation of heterogeneity. We have developed a set of heterogeneity indices (HI's) that can be used to guide decisions. Examples from drug discovery and diagnostics will be discussed.

11:25 Integrating Novel Technologies to Identify Small Molecules that Drive Translational Research and Therapeutics

MichellePalmerMichelle Palmer, Ph.D., Co-Director, Center for the Development of Therapeutics, Broad Institute of Harvard and MIT

Advances in human genetics have lead to new drug discovery strategies that may lower the rate of attrition when translated to human trials. Molecular characterization of patient-derived samples is providing new insights into the root cause of many diseases. Many of these insights point to targets that have traditionally been challenging for small-molecule therapeutics. Identification of drugs to modulate targets where knowledge of the function in disease is poorly understood and processes such as disruption of these novel targets require innovation in chemistry, phenotypic cell-based assays and target identification studies. At the Broad Institute, we have integrated technology across all aspects of lead identification in an effort to realize the benefit of the genes to drugs approach in multiple disease areas including psychiatric disease. Examples that illustrate the application of novel targets and pathways in a phenotypic screening approach will be presented.

12:00 pm Luncheon Presentation: Phenotypic Screening of Compounds and Drug Combinations in BioMAP® Oncology Systems Reveals Divergent Effects on Clinically-Relevant Biomarkers 

Pattison_ScottScott Pattison, Ph.D., Director, Business Development, DiscoveRx, BioSeek Division

BioMAP® Oncology Systems model host-tumor microenvironments using human primary fibroblasts or endothelial cells co-cultured with PBMC and select cancer cell lines. The signaling environment modeled therein significantly impacts drug activity on clinically relevant biomarkers. Consistent with clinical reports, we detect enhanced activities with the combination of Dabrafenib (BRAF inhibitor) and Trametinib (MEK inhibitor) as compared to the respective monotherapies. BioMAP® oncology systems provide a predictive human model to assess therapeutic strategies prior to clinical testing.

High-Content Analysis

1:30 Chairperson’s Remarks

Anne E. Carpenter, Ph.D., Broad Institute of Harvard and MIT

1:35 Advancing Drug Discovery through the Application of High-Content Phenotypic Profiling

BeverleyIsherwoodBeverley Isherwood, Ph.D., Team Leader, AstraZeneca R&D

High-content analysis (HCA) has developed into an important platform for drug discovery. We describe approaches taken at AstraZeneca to develop and implement multiparametric high-content assay panels in physiologically-relevant models of disease to identify and validate targets, predict toxic liability and characterize compound mode of action. We discuss the tools and workflows adopted to facilitate the efficient, consistent and reliable application of HCA at multiple stages of drug discovery.

2:00 High-Content Imaging Workflows for Screening and Assay Development in Oncology Drug Discovery

Joern Hopke, Ph.D., Senior Research Investigator, Sanofi

High-content imaging and phenotypic screening create distinctive challenges for data management and subsequent data analysis. We have assembled an array of data management, processing and analysis tools into efficient and flexible workflows for large-scale HCS and small-scale assay development alike. Examples of a screening campaign with successive multivariate data analysis for compound MOA deconvolution and a number of phenotypic 3D assays relevant to oncology drug discovery will be presented.

2:25 Advancing Phenotypic Drug Discovery in Cancer through Combined High-Content Imaging and Reverse Phase Protein Array Technology

NeilCarragherNeil Carragher, Ph.D., Principal Investigator, Edinburgh Cancer Research UK Centre, University of Edinburgh

Limited understanding of drug mechanism-of-action and pharmacological resistance contributes to poor efficacy and attrition at later stages of drug discovery and development. We demonstrate how multiparametric high-content imaging and Reverse Phase Protein Microarray (RPPA) technologies can combine to enable a robust and unbiased approach to profiling compound mechanism-of-action and optimizing efficacy within complex in vitro and in vivo settings. We will provide case studies demonstrating how we have applied these technologies to advanced models of cancer to progress new chemical entities identified initially as hits from phenotypic screens towards late-stage preclinical development.

 Propagenix2:50 Conditionally Reprogrammed (CR) Cell Technology: A Cell Culture Platform Delivering a Continuous Source of Personalized Primary Epithelial Cells and Tumor Cells for Drug Discovery
Pollok_BrianBrian Pollok, Ph.D., President & CEO, Propagenix Inc
CR cell technology is a validated method for establishing primary epithelial and tumor cell cultures of unlimited propagation, with the ability for forward engineering of patient-specific primary cells. Most importantly, CR-cultured primary cells readily return to their normal differentiated phenotype upon removal of the cells from the CR culture media.

3:05 Refreshment Break in the Exhibit Hall with Poster Viewing

3:45 Discovering Unexpected Phenotypes Using Image-Based Profiling

AnneCarpenterAnne E. Carpenter, Ph.D., Director, Imaging Platform, Broad Institute of Harvard and MIT

Microscopy images contain rich information about the state of cells, tissues and organisms. Our laboratory is extracting patterns of morphological perturbations (“signatures”) from images in order to identify similarities between various chemical or genetic treatments. Our goal is to classify drug mechanisms of efficacy and toxicity, distinguish cancer-relevant proteins, and identify biomarkers of disease. We hope to make perturbations in cell morphology as computable as other large-scale functional genomics data.

4:15 Advances in Phenotypic HTS Reporter Assays and Opportunities for Underserved Disease

James Inglese, Ph.D., National Center for Advancing Translational Sciences, National Institutes of Health

Emerging sophistication in assays developed to model pathogenic pathways has the potential to reveal new insights regarding possible routes to pharmacological intervention and in some cases directly to clinical candidates from the repositioning of approved drugs. Specific examples from my lab at NIH focused on assay development and high-throughput screening will be discussed to illuminate these ideas in the context of disease foundation-sponsored collaborative partnerships.

4:45 The Implementation of Phenotypic Cell-Based Assays and High-Content Imaging in Translational Research: Choosing the Right Tools for the Job

AnthonyDaviesAnthony M. Davies, Ph.D., Center Director, Translational Cell Imaging Queensland (TCIQ), Institute of Health Biomedical Innovation, Queensland University of Technology


5:15 Welcome Reception in the Exhibit Hall with Poster Viewing

6:15 End of Day One/Short Course Registration

Recommended DINNER Short Course*

6:30-9:30 (SC3) Stem Cell Models for Drug Discovery 

*Separate registration required.

Tuesday, November 18

Complex Physiologically-Relevant Cellular Models for Phenotypic Drug Discovery

8:30 Chairperson’s Remarks

Michael Jackson, Ph.D., Sanford-Burnham Medical Research Institute

8:35 Rapid Target Screening Technology for ‘Hit-Prioritization’ and ‘New Target Identification’ in Phenotypic Screening Scenario

Saxena_ChaitanyaChaitanya Saxena, Ph.D., CEO, Shantani

An affinity chromotagraphy method that can utilize unmodified ‘bait’ molecule for Target-identification can rapidly screen the targets of multiple ‘Hits’ coming out of phenotypic screens. We will present our technology platform that primarily relies on deconvoluting the targets of unmodified phenotypically screened ‘Hits’ in medium throughput fashion. Utility of Shantani’s target identification technology platforms in Phenotypic Screening Scenario for ‘Hit-Prioritization’ and ‘New Target Identification’ will be presented.

9:05 Use of Patient-Derived Cells in Disease-in-a-Dish Phenotypic Screens

Michael Jackson, Ph.D., Senior Vice President, Drug Discovery and Development, Conrad Prebys Center for Chemical Genomics, Sanford-Burnham Medical Research Institute

Reprogramming of somatic cells to induced pluripotent stem cells (iPSC) has enabled a new paradigm in drug discovery that leverages disease-in-a-dish phenotypic screens. Technical challenges in the scalability and reproducibility of assays based on iPS (patient) cells remain a major hurdle in executing such screens, especially where mature differentiated cell types are required. Progress on the use of this approach to identify drugs (repurposed) to treat genetic disease will be presented.

9:30 Phenotypic Screening for Alzheimer’s Disease Drug Discovery in Stem Cell-Derived and Primary Brain Cells

KimTaeWanTae-Wan Kim, Ph.D., Associate Professor, Pathology and Cell Biology, Columbia University Medical Center

The development of effective and safe therapeutics for complex neurodegenerative diseases, such as Alzheimer’s disease (AD) is hampered in part by lack of physiological cell models. We will discuss our progress on phenotypic, high-throughput screening platforms targeting key AD-relevant cellular pathways (i.e. tau, amyloid and apoE) in primary and stem cell-derived neurons and glial cells. Our approach will facilitate preclinical discovery of promising therapeutic lead molecules for AD drug discovery.

10:05 Coffee Break in the Exhibit Hall with Poster Viewing

10:45 Human Bronchial Epithelial Cells as a Phenotypic Model in Cystic Fibrosis Drug Discovery

TimYoungTim Young, Ph.D., Research Fellow, Vertex Pharmaceuticals

Cystic fibrosis (CF) is caused by mutations in the CFTR gene, which result in decreased ion and fluid transport across bronchial epithelium. Human bronchial epithelial (HBE) cells grown in culture at an air-liquid interface retain the phenotype of airway epithelium with respect to ion and fluid transport, and mucociliary biology. HBEs derived from subjects with CF are impaired in these functions and offer a phenotypic model to develop CFTR modulators.

11:10 Zebrafish Phenotypic Discovery for Small Molecules that Augment Kidney Regeneration

AndreasVogtAndreas Vogt, Ph.D., Associate Professor, Computational and Systems Biology, University of Pittsburgh

Kidney regeneration after acute injury (AKI) is limited by formation of fibrotic scar tissue. Through zebrafish phenotypic screening we identified small molecules that prevent scarring and augment kidney regeneration after injury. Transcriptional profiling suggested a cellular mechanism involving aberrant cell cycle progression, which was confirmed experimentally in zebrafish and mice. The data illustrate the utility of a zebrafish quantitative systems pharmacology approach to discover new potential treatments for AKI.

11:35 Cancer Drugs on the Fly: Whole-Animal Chemical Screening in Drosophila Identifies Drug Interactions with Stem Cells, Stem Cell Tumors and the Stem Cell Microenvironment

Michele Markstein, Ph.D., Assistant Professor, Biology, UMass Amherst

Here we report the development of an in vivo chemical screening platform using tumor models in the adult Drosophila intestine to study the interaction of stem cells with drugs. Strikingly, we find that some FDA-approved chemotherapeutics that can inhibit the growth of Drosophila stem cell tumors can paradoxically promote the hyper-proliferation of their wild type counterparts. These results reveal an unanticipated side effect on stem cells that may drive tumor recurrence. We show that this side effect is mediated by the stem cell microenvironment, which we demonstrate responds to a wide spectrum of chemotherapy drugs. We propose that the same side effect may occur in humans based on our finding that it is driven by the evolutionarily conserved JAK-STAT pathway. To identify additional compounds that can inhibit tumors without inducing side effects on the stem cell microenvironment, we screened a library of 6,100 small molecules, from which we report the identification of 10 compounds that inhibit tumors without inducing the growth promoting side effect. Altogether, our results highlight the importance of using in vivo models to study the effects of drugs on stem cells: we show that the impact of a chemotherapy drug on the stem cell microenvironment is just as important as its impact on the stem cell itself.

12:00 pm Close of Phenotypic Drug Discovery - Part I Conference

Tuesday, November 18

12:00 pm Conference Registration

Phenotypic Screening of 3D Models

1:30 Chairperson’s Opening Remarks

Aron Jaffe, Ph.D., Novartis Institutes for BioMedical Research

1:35 Developing and Utilizing 3D Culture Systems for Novel Target Discovery

AronJaffeAron Jaffe, Ph.D., Senior Investigator, Developmental and Molecular Pathways, Novartis Institutes for BioMedical Research

Target identification and validation have historically relied on immortalized or tumor cell lines grown on plastic. Recent techniques involving growth of cells in three-dimensional matrices have enabled modeling of cellular processes in an environment that more closely resembles the in vivo setting. This presentation will highlight the strategies for designing complex cellular assays for target discovery using medium and high-throughput screening methods in three dimensions.

2:00 3D versus 2D: Insight from Pharmacology and Genomics

Jean-Louis Klein, Ph.D., Target and Pathway Validation, Platform Technology and Science, GlaxoSmithKline

2:30 Technology Showcase (Sponsorship Opportunities Available)

3D Cellular Models for Drug and
Target Screening

This showcase provides an opportunity for sponsoring companies to showcase their new and emerging 3D cell models and technologies for the next generation of drug and target phenotypic screening.

2:30 High-Throughput Compatible Co-Spheroid Model Analyzing Compound Effects on Both Tumor and Stroma Cells

Jan E. Ehlert, Ph.D., Head, Cellular Drug Discovery, ProQinase GmbH

A spheroid-based co-culture system for the simultaneous analysis of compound effects on the proliferation of tumor as well as of stroma cells was established. The modular HTS-compatible system reveals results reflecting cell-specific drug susceptibility and cell/cell interactions.

2:50 High Throughput Organ-on-a-Chip Models for Predictive Toxicology and Efficacy Testing

Joore_JosJos Joore, Ph.D., Chief Business Officer, MIMETAS BV

OrganoPlates™ are a novel microfluidic culture platform enabling long-term, membrane-free 3D co-culture models in a microtiterplate format. We have developed a large variety of tissue- and disease models, applicable for drug testing and evaluation. The platform is compatible with standard readout equipment, making the technology suitable for high‐throughput automation.

3:10 3D InSight™ Microtissues for Drug Discovery and Development

Kelm_JensJens M. Kelm, Ph.D., CSO, Co. Founder, InSphero AG

Increasing the biological significance of in vitro models to better de-risk drug failure will foster the drug discovery and developmental process. Microtissue models tailored either for efficacy testing or safety testing have shown to be a versatile culture format to be used throughout the whole development process.

3:30 Refreshment Break in the Exhibit Hall with Poster Viewing

High-Content Analysis of Tumor Spheroid Models

4:25 Chairperson’s Remarks

David Nolte, Ph.D., Professor, Physics, Purdue University; President, Animated Dynamics, Inc.

4:30 Drug Discovery and Development of Novel Anticancer Agents: Applications of Novel 3D Multicellular Tumor Spheroid Models

DanielLaBarberaDaniel V. LaBarbera, Ph.D., Assistant Professor, Drug Discovery and Medicinal Chemistry, The Skaggs School of Pharmacy and Pharmaceutical Sciences, The University of Colorado

Aberrant regulation of epithelial-mesenchymal transition (EMT) is a driving force in the most prominent human diseases. In particular, EMT-driven tumor progression promotes the expansion of cancer stem cells, drug resistance, and the mesenchymal phenotype, which is invasive with a high metastatic potential. Therefore, one therapeutic strategy to prevent metastatic dissemination is to develop small molecule drugs that can revert the mesenchymal phenotype to the more benign epithelial state. Using novel 3D multicellular tumor spheroid (MCTS) models of EMT, suitable for high-throughput and high-content screening (HTS/HCS), we have identified lead compounds that block TCF-transcription, which regulates the mesenchymal phenotype in colorectal cancer.

4:55 Novel Stromal Targets that Support Tumor Spheroid Formation

ShaneHormanShane R. Horman, Ph.D., Research Investigator, Advanced Assay Group, Genomics Institute of the Novartis Research Foundation

The stroma of solid organ tumors influences all phases of tumor progression indicating that environment may be dominant to the genetics of cancer cells. To expand these concepts to early stage drug discovery we have scrutinized the interactions between colon stroma and colorectal carcinoma (CRC) cells in a high-content co-culture 3D spheroid screen. Subsequently we were able to identify novel fibroblast genes that, when depleted, abrogate CRC spheroid formation revealing an extended chemotherapeutic space by which to target tumors.

5:20 Developing Biodynamic Screening Assays for 3D Live-Tissue Models

DavidNolteDavid Nolte, Ph.D., Professor, Physics, Purdue University; President, Animated Dynamics, Inc.

Biodynamic screening performs three-dimensional functional imaging of living tissue by measuring drug-induced changes in intracellular dynamics. It is compatible with many 3D tissue formats, including tumor spheroids grown in bioreactors or in multiwell plates, as well as tissue biopsies and other organotypic models. Dynamics-based phenotypic profiling of tissues provides a new type of high-content screening. This talk presents 3D assays being developed for chemosensitivity and resistance, proliferation and toxicity screening.

5:45 Close of Day

5:45 Short Course Registration

Recommended Dinner Short Course*

6:00-9:00 (SC5) Expert ThinkTank: How to Meet the Need for Physiologically-Relevant Assays? 

*Separate registration required.

Wednesday, November 19

7:30 am Registration and Morning Coffee

Phenotypic Data Analysis and Modeling

8:00 Chairperson’s Remarks

Peter Sorger, Ph.D., Otto Krayer Professor, Systems Biology, Harvard Medical School; Visiting Professor, Biological Engineering, MIT

8:10 Cellular Assays for Cancer Pharmacology

Peter Sorger, Ph.D., Otto Krayer Professor, Systems Biology, Harvard Medical School; Visiting Professor, Biological Engineering, MIT

A major goal of the Harvard Medical School Library of Integrated Network-based Cellular Signatures (LINCS) Center is to collect and disseminate the data and analytical tools needed to understand how human cells respond to perturbations created by exposure to drugs. Our center has developed assays to systematically query cell responses to kinase inhibitors and tools to analyze and visualize the results.


8:35 In silico Lead Finding through Holistic Understanding of Screening Data from Multiple Approaches

Meir GlickMeir Glick, Ph.D., Head, In Silico Lead Discovery – Cambridge, Center for Proteomic Chemistry, Novartis Institutes for BioMedical Research

The changing drug discovery environment presents a richer, more complicated and novel data landscape. How can state of the art data analytics increase the probability of a lead compound to be disease relevant? We will discuss how in silico approaches actively shape the lead discovery process: informing on relevant assays, compounds subset design to probe the biology, visualization of complex biological data, models elucidating target/MOA hypothesis and design of chemical matter.

9:00 Genome-Wide RNAi-Compound Epistasis to Elucidate Drug Pathways

JeremyJenkinsJeremy L. Jenkins, Ph.D., Senior Investigator, Developmental and Molecular Pathways, High-Throughput Biology, Novartis Institutes for BioMedical Research

Following phenotypic compound screening, target elucidation remains a bottleneck for follow-up. We investigate examples of genome-wide siRNA screening in the presence of sensitizing compound doses in order to find nodes that sensitize or suppress drug activity and map ‘compound-effect’ pathways. Epistatic shifts in drug activity are observed by intended target knockdown as well as siRNA off-target knockdown caused by miRNA-like polypharmacology of unintended siRNA seed sequence binding to mRNA 3’UTRs.

9:25 Towards Open Source Software for High-Content Screening and Phenotypic Data Analysis

Frans Cornelissen, Principal Scientist, Translational Informatics, Janssen Pharmaceutical Companies of Johnson & Johnson

A crucial step towards efficient mining of HCS data is the availability of a robust platform for HCS data management and (semi-)automatic cellular phenotype identification and classification. However, for many small and medium-size screening groups, commercial software may be too expensive. We are in the process of making our internally developed Phaedra software available as an open source product. Examples will be presented to show that Phaedra is a sound foundation that can be used as an integrated pattern recognition and machine learning environment for HCS.

9:50 Coffee Break in the Exhibit Hall with Poster Viewing

10:45 Chairperson’s Remarks

Rosemarie Hunziker, Ph.D., Director, Tissue Engineering and Regenerative Medicine, National Institute of Biomedical Imaging and Bioengineering, National Institutes of Health

10:50 Automated Learning of Perturbation Models

Armaghan Naik, Ph.D., Lane Fellow, Lane Center for Computational Biology, Carnegie Mellon University

Traditional HCS methods detect perturbed phenotypes but do not allow identification of the specific changes in cell organization underlying them; results also cannot be easily compared across different HCS systems. We developed methods to convert images into models of cell components and how perturbagens affect them, and have also demonstrated that active machine learning can build models of the effects of many drugs on many targets without exhaustive experimentation.

11:15 An Evolutionary Dynamics Approach to Studying Tumor Resistance

Arijit Chakravarty, Ph.D., Director, Modeling and Simulation, Takeda Pharmaceuticals

11:40 pm Close of Phenotypic Drug Discovery - Part II Conference