LARA MANGRAVITE SAGE BIONETWORKS ON BEHALF OF THE RA CHALLENGE ORGANIZING TEAM The DREAM Rheumatoid Arthritis Responder Challenge: Motivation, Data, Scoring.

Slides:



Advertisements
Similar presentations
Regulation of Consumer Tests in California AAAS Meeting June 1-2, 2009 Beatrice OKeefe Acting Chief, Laboratory Field Services California Department of.
Advertisements

Taking Research and Development to the Clinic: Issues for Physicians AAAS/FDLI Colloquium I Diagnostics and Diagnoses Paths to Personalized Medicine Howard.
CZ5225 Methods in Computational Biology Lecture 9: Pharmacogenetics and individual variation of drug response CZ5225 Methods in Computational Biology.
Introduction to Challenge 2 The NIEHS - NCATS - UNC DREAM Toxicogenetics Challenge THE DATA Fred A. Wright, Ph.D. Professor and Director of the Bioinformatics.
Genetic Analysis in Human Disease
Clinical Trial Designs for the Evaluation of Prognostic & Predictive Classifiers Richard Simon, D.Sc. Chief, Biometric Research Branch National Cancer.
Antibody Response During a PRRS Outbreak can be Predicted Using High-Density SNP Genotypes Nick V.L. Serão 1 *, R.A. Kemp 2, B.E. Mote 3, J.C.S. Harding.
Perspectives from Human Studies and Low Density Chip Jeffrey R. O’Connell University of Maryland School of Medicine October 28, 2008.
Breast cancer is a complex and heterogeneous disease Tumor samples Protein expression Clinical features Mutational status Adapted from TCGA, Nature 2012.
Recursive Partitioning Method on Survival Outcomes for Personalized Medicine 2nd International Conference on Predictive, Preventive and Personalized Medicine.
Journal Club Alcohol, Other Drugs, and Health: Current Evidence May–June 2013.
1 FSTL4 and SEMA5A are associated with alcohol dependence: meta- analysis of two genome-wide association studies Kesheng Wang, PhD Department of Biostatistics.
Model and Variable Selections for Personalized Medicine Lu Tian (Northwestern University) Hajime Uno (Kitasato University) Tianxi Cai, Els Goetghebeur,
The Pursuit of Better Medicines through Genetic Research Terri Arledge, DVM US Department Head Drug Development Genetics.
Methods to Analyse The Economic Benefits of a Pharmacogenetic (PGt) Test to Predict Response to Biologic Therapy in Rheumatoid Arthritis, and to Prioritise.
The Cost-Effectiveness and Value of Information Associated with Biologic Drugs for the Treatment of Psoriatic Arthritis Y Bravo Vergel, N Hawkins, C Asseburg,
Journal Club Alcohol, Other Drugs, and Health: Current Evidence March–April 2015.
Give me your DNA and I tell you where you come from - and maybe more! Lausanne, Genopode 21 April 2010 Sven Bergmann University of Lausanne & Swiss Institute.
Correlational Designs
Identifying RA patients from the electronic medical records at Partners HealthCare Robert Plenge, M.D., Ph.D. VA Hospital July 20, 2010 HARVARD MEDICAL.
Host Genomics in WIHS  The WIHS GWAS data set  Concept Sheet  Data use agreement  Data transfer  Analytic support.
Elizabeth Karlson, MD Associate Professor of Medicine
A crowdsourcing effort that poses questions (Challenges) about biology, modeling and data analysis: – Transcriptional networks.
Crowdsourcing pharmacogenomic data analysis: PGRN-Sage RA Responder Challenge PGRN Spring Meeting April 30, 2013 HARVARD MEDICAL SCHOOL.
Biomarker and Classifier Selection in Diverse Genetic Datasets J AMES L INDSAY 1 E D H EMPHILL 2 C HIH L EE 1 I ON M ANDOIU 1 C RAIG N ELSON 2 U NIVERSITY.
Sequencing TRAF1 in patients with rheumatoid arthritis Bruce C. Jobse Medical and Population Genetics Broad Institute.
A centre of expertise in digital information management UKOLN is supported by: Monica Duke Project.
Journal Club Alcohol, Other Drugs, and Health: Current Evidence May–June 2012.
University of Washington Institute of Technology Tacoma, WA, USA Ecole des Hautes Etudes en Santé Publique Département Infobiostat Rennes, France Isabelle.
Precision Medicine A New Initiative. The Concept of Precision Medicine (PM) The prevention and treatment strategies that take individual variability into.
Sage Bionetworks A non-profit organization with a vision to enable networked team approaches to building better models of disease BIOMEDICINE INFORMATION.
L 1 Chapter 12 Correlational Designs EDUC 640 Dr. William M. Bauer.
Biostatistics Case Studies 2007 Peter D. Christenson Biostatistician Session 3: Incomplete Data in Longitudinal Studies.
The Early Days of an Investigator in WIHS: Grants and Projects By Bani Tamraz, Pharm.D., Ph.D. Associate Clinical Professor School of Pharmacy.
PROTOCOL 588: OVERALL ASSESSMENT Concept is sound, well supported Vector is safe Transgene product is reasonably safe Overall, a phase I study seems acceptable.
The Broad Institute of MIT and Harvard Classification / Prediction.
Pharmacogenetics & Pharmacogenomics Personalized Medicine.
Wang Y 1,2, Damaraju S 1,3,4, Cass CE 1,3,4, Murray D 3,4, Fallone G 3,4, Parliament M 3,4 and Greiner R 1,2 PolyomX Program 1, Department.
Journal Club Alcohol, Other Drugs, and Health: Current Evidence May–June 2014.
Genome-Wide Association Study (GWAS)
Statistical Review: Recursive Partitioning Identifies Patients at High and Low Risk for Ipsilateral Tumor Recurrence After Breast- Conserving Surgery and.
Personalized Medicine Dr. M. Jawad Hassan. Personalized Medicine Human Genome and SNPs What is personalized medicine? Pharmacogenetics Case study – warfarin.
The Use of Predictive Biomarkers in Clinical Trial Design Richard Simon, D.Sc. Chief, Biometric Research Branch National Cancer Institute
Melanie Dunn 9/20/2011.  Rheumatoid arthritis (RA) is a chronic disease that leads to inflammation of the joints and surrounding tissues that can also.
Sage Bionetworks A non-profit organization with a vision to enable networked team approaches to building better models of disease BIOMEDICINE INFORMATION.
BioPaths-Catalyze Drug Discovery, Development and Clinical Research
Using Predictive Classifiers in the Design of Phase III Clinical Trials Richard Simon, D.Sc. Chief, Biometric Research Branch National Cancer Institute.
Risk Prediction of Complex Disease David Evans. Genetic Testing and Personalized Medicine Is this possible also in complex diseases? Predictive testing.
AISTATS 2010 Active Learning Challenge: A Fast Active Learning Algorithm Based on Parzen Window Classification L.Lan, H.Shi, Z.Wang, S.Vucetic Temple.
Utility of Genotyping in Pharmaceutical Target (gene) Discovery and Drug Response Anne Westcott EST-Informatics.
D4FF55A0-6B6F BF422A9BA9 Present by: Xiao Chen On December 7, 2015.
Using State Tests to Measure Student Achievement in Large-Scale Randomized Experiments IES Research Conference June 28 th, 2010 Marie-Andrée Somers (Presenter)
Approaches to quantitative data analysis Lara Traeger, PhD Methods in Supportive Oncology Research.
Understanding GWAS SNPs Xiaole Shirley Liu Stat 115/215.
Docker in Open Science Data Analysis Challenges Bruce Hoff Principal Software Engineer, Sage Bionetworks.
Kelci J. Miclaus, PhD Advanced Analytics R&D Manager JMP Life Sciences
Changing the trajectory of drug R&D
Acceptable changes in quality attributes of glycosylated biopharmaceuticals
Changing the trajectory of drug R&D
5 Different Observational Datasets: Pros & Cons
Classification with Gene Expression Data
Gene Hunting: Design and statistics
What Is a Biosimilar?. Biosimilar Application in RA Clinical Practice: Current Knowledge and New Data  
Model Enhanced Classification of Serious Adverse Events
Independent baseline predictors of non-remission at 24 months of follow-up in the SWEFOT trial population. Independent baseline predictors of non-remission.
New approaches to personalized medicine for asthma: Where are we?
Satisfaction with control of RA
Nadia Howlader, PhD National Cancer Institute
Incidence rates of hospitalised infection per 100 person-years, standardised for age and sex, among patients with RA from five RA registries and one RA.
Hong Zhang, Judong Shen & Devan V. Mehrotra
Presentation transcript:

LARA MANGRAVITE SAGE BIONETWORKS ON BEHALF OF THE RA CHALLENGE ORGANIZING TEAM The DREAM Rheumatoid Arthritis Responder Challenge: Motivation, Data, Scoring and Results

Challenge Organizers Eli Stahl, Mt Sinai Gaurav Pandey, Mt Sinai Jing Cui, Brigham and Women’s Andre Falcao, U Lisbon Robert Plenge, Merck Peter Gregersen, Feinstein Institute Jeff Greenberg, Corrona Dimitrios Pappas, Corrona Kaleb Michaud, Arthritis Internet Registry Generators of Training Dataset Solly Sieberts Abhi Pratap Christine Suver Bruce Hoff Thea Norman Venkat Balagurusamy Stephen Friend Gustavo Stolovitzky Funders

~30% of RA patients fail to respond to anti-TNF therapy Rheumatoid Arthritis Treatment -- Predicting nonresponse would assist in precision medicine, clinical trial design, and development of new therapies Robert Plenge

n=2,706 Pharmacogenetics of antiTNF response DrugN SNP- heritability (se) P-value All patients (0.10)0.02 etanercept7160 (0.34)0.5 infliximab (0.29)0.02 adalimumab (0.25)0.08 infliximab + adalimumab (0.13)0.003 Ciu and Stahl et al PLoS Genetis 2013 Eli Stahl

Rationale Given sizable estimated heritability, is it possible to use genetic features to predict treatment response?  Polygenic approach: Combined influence of weak effects  Population subtypes: Not all individuals react similarly  Does genetic heritability foretell genetic prediction?

RA Responder Challenge Design Discovery (phase I) GWAS of treatment response in RA (n≈2,700 patients) GWAS of treatment response in RA (n≈2,700 patients) Genomic data (e.g., expression profiling) Genomic data (e.g., expression profiling) Polygenic SNP predictor of response Refine model Plenge et. al. Nature Genetics 2013

Discovery (phase I) Validation (phase II) GWAS of treatment response in RA (n≈2,700 patients) GWAS of treatment response in RA (n≈2,700 patients) Genomic data (e.g., expression profiling) Genomic data (e.g., expression profiling) Polygenic SNP predictor of response Refine model Submit models GWAS of treatment response in RA (n≈1,100 patients) GWAS of treatment response in RA (n≈1,100 patients) Score models RA Responder Challenge Design Plenge et. al. Nature Genetics 2013

Discovery (phase I) Validation (phase II) GWAS of treatment response in RA (n≈2,700 patients) GWAS of treatment response in RA (n≈2,700 patients) Genomic data (e.g., expression profiling) Genomic data (e.g., expression profiling) Polygenic SNP predictor of response Refine model GWAS of treatment response in RA (n≈1,100 patients) GWAS of treatment response in RA (n≈1,100 patients) Submit models Score models RA Responder Challenge Design Plenge et. al. Nature Genetics 2013

RA Challenge Data Genotypes ~ 2.3 million SNPs Clinical ~ 6 traits N=2076Response Discovery Dataset Combine set from 4 studies Test Data Genotypes ~ 2.3 million SNPs Clinical ~ 6 traits N=723 Generated for this challenge

RA Challenge: Build the best possible predictors of anti-TNF  response in RA TEAM PHASE February - June 2014 Self-aggregate into teams and build the best possible predictor of response. COMMUNITY PHASE July - October 2014 Work together across teams to assess the contribution of genetics to prediction. Team Phase Community Phase

RA Responders Challenge Predict treatment response as measured by change in disease activity score (DAS28) in response to ant- TNF  therapy.  Scoring: Average rank of pearson correlation and spearman correlation. Identify poor responders to anti-TNF  therapy as defined by EULAR criteria.  Scoring: Average rank of AUC and PR.

Team Phase Results Subchallenge 1: Predicting deltaDAS Subchallenge 2: Predicting nonresponders Best models: Team Guan Lab & Team SBI_Lab Solly Sieberts 32 teams

The Community Phase (July – October) Work in collaboration to determine: -- Whether genetic information contributes in a meaningful way to predictions? -- Best possible predictors of response. -- What components of the modeling approaches are most beneficial for this question.

Community Phase Participants

Community Phase Logistics First part: teams split into groups and shared knowledge to help inform one another’s efforts Second part: all teams came together to devise an analytical plan to explicitly address these questions.

Teams share ideas and then work individually to provide: Do models using genetic features improve on prediction relative to clinical models? What is the contribution of feature selection vs. modeling algorithm on performance? Does the use of biological priors in feature selection improve relative to random selection? Can supervised ensemble approach improve upon individual predictions?

Subchallenge 1:Predicting deltaDAS

Subchallenge 2: Predicting Nonresponders

Ensemble Modeling by Gaurav Pandey

Conclusions Gaussian Process Regression appears to work best with this type of problem. SNP selection more important than algorithmic selection in most cases. Genetic information improves prediction of nonresponders over use of clinical information. Ability to predict response based on clinical features may be valuable to clinicians in and of themselves.

Today’s Speakers: Best Performers from Independent Team Phase Fan Zhu on behalf of Team Guan Lab  A generic method for predicting clinical outcomes and drug response Javier Garcia-Garcia on behalf of Team SBI_Lab  Predicting response to arthritis treatments: regression-based gaussian processes on small sets of SNPs