STEERING COMMITTEE MEETING February 20 2020 TollFree 1

STEERING COMMITTEE MEETING February 20, 2020 Toll-Free: 1 -866 -901 -6455 ● Long Distance: 1 -562 -247 -8421 Access Code: 888 -196 -492 https: //attendee. gotowebinar. com/register/5392020819915708941 ***************************** February 2020 Final deliverables 0

Final Deliverables Discussion • Describe accomplishments and work remaining for each of the seven network milestones and workgroups • Identify timelines and groups associated with remaining work • Outline plan for remaining time in e. MERGE III and close out call June 4 th, 2020 from 2: 00 -5: 00 pm EST. February 2020 Final deliverables 1

Milestone 1: Clinical Annotation Conduct penetrance analysis in conditions with sufficient data and assess impact on clinical outcomes; describe lessons learned. Accomplishments: • Developed outcomes forms for 10 disease areas • Data was insufficient for 4 disease areas Sufficient Data Positive Subjects Forms Collected Arrhythmias 172 73 34 28 203 113 Breast Cancer 397 137 Cardiomyopathy 239 114 Colorectal Cancer 235 206 Ehlers-Danlos syndrome 4 2 Chronic Kidney Disease 14 13 Ornithine transcarbamylase deficiency 1 7 Tuberous Sclerosis 16 1 Aortic Dilation/ Aortopathy Familial Hypercholesterolemia • Analyses were performed for 6 disease areas with Insufficient Data sufficient data including arrhythmias, aortic disease, cardiomyopathy, FH, breast cancer and colon cancer February 2020 Final deliverables 2

Milestone 1: Clinical Annotation Conduct penetrance analysis in conditions with sufficient data and assess impact on clinical outcomes; describe lessons learned. Remaining work: Arrhythmias (Dan Roden): Project 1: Code-based > stand-alone paper Project 2: ECG based (need QT intervals from network before end of March) > may combine with Project 1 publication Project 3: Brugada syndrome FH (Iftikhar Kullo): Profile in large cohort (public health perspective; combined penetrance, risk): MCS circulated Request for each site to get additional data (or minor clarification) on 120 genotype+ subjects Breast cancer (Wendy Chung): Redoing analysis with indications for referrals. Need outstanding forms completed from sites Aortopathy (Iftikhar Kullo): Will give guidance for metrics needed to calculate z scores February 2020 Final deliverables 3

Milestone 1: Clinical Annotation Penetrance Lessons learned • How do we study penetrance? • In people not ascertained by phenotype or genotype. • How people were ascertained was not straightforward, e. g. in cardiology clinic the ascertainment may or may not have been relevant to lipids • In a large enough sample size (E 3 OK for BRCA 1/2, LDLR/APOB, Arrhythmia). • Penetrance would require a lot of data for rare disorders or risk alleles (OR in larger dataset easier). • Conflation of pathogenicity classification vs. penetrance (also, are LPs true Ps? ) • Need specific outcomes forms (some of which can be completed by participants) or records (such as ECG) and detailed age of onset or screening data. • Forms need to be tested and time to review charts must be supported • This represents an iterative process, in that, while testing is important to make sure the forms work to collect data, we’ve had several instances where the data collected based on the initial form construction has some deficiencies or gaps that led to requests for second review. This should be accounted for. • In some cases where clinical diagnosis may not be made despite presence of disease, further phenotyping was needed (tuberous sclerosis, NF 2) for both variant classification and penetrance assessment. February 2020 Final deliverables 4

Milestone 2: Outcomes accomplishments Determine the impact of return of genetic results (Ro. R) on patients’ immediate outcomes, 6 months and when available 12 months after Ro. R in conditions with sufficient data. • 15 outcomes forms developed, and data harmonized • Two data freezes for interim analysis disseminated to Network • Data entry issues identified, abstraction guides developed • Additional data needed for final assessment identified • 6 -month outcomes assessment in final stages, final data freeze in March for remaining data issues. February 2020 Final deliverables 5

Milestone 2: Outcomes remaining work Determine the impact of return of genetic results (Ro. R) on patients’ immediate outcomes, 6 months and when available 12 months after Ro. R in conditions with sufficient data. • Complete final six-month analysis • Needs: Data freeze three (March 18 th, 2020) • Prepare and submit Outcomes manuscripts (on going) • NT 296 Peterson et al. Collection and analysis of large-scale outcome measures following targeted next generation sequencing • Needs: ICD & CPT data refresh • Complete 12 -month outcomes data entry • Focus on high impact phenotypes • Needs: Data freeze three (March 18 th, 2020) February 2020 Final deliverables 6

Milestone 2: Ro. R Determine the impact of return of genetic results (Ro. R) on participants’ immediate outcomes, 6 months and when available 12 months after Ro. R in conditions with sufficient data. Accomplishments: • Participant surveys (Participant survey subgroup of Ro. R) • Coordinated across sites at baseline (decisional conflict only) and 1 - and 6 -12 -months post-disclosure • Data dictionaries reconciled for most questions (Wynn & Milo Rasouly), data being placed in one REDCap database at CC. • NT 373: Psychologic impact on participants of receiving positive genomic results in e. MERGE III – work in progress. • NT 363: Participants perceived clinical and personal utility of receiving positive genomic results in e. MERGE III – to follow NT 373. February 2020 Final deliverables 7

Milestone 2: Ro. R Determine the impact of return of genetic results (Ro. R) on participants’ immediate outcomes, 6 months and when available 12 months after Ro. R in conditions with sufficient data. Remaining work: • Participant surveys (Participant survey subgroup of Ro. R) • Finish reconciling data dictionaries, data being placed in one REDCap database at CC. (4/1/2020) • NT 373: Psychologic impact on participants of receiving positive genomic results in e. MERGE III – finish analysis and manuscript. (6/1/2020) • NT 363: Participants perceived clinical and personal utility of receiving positive genomic results in e. MERGE III – analysis and manuscript. (7/1/2020) February 2020 Final deliverables 8

Milestone 3: EHRI Improve and/or standardize genomic clinical decision support (CDS) for return of clinically relevant genetic or incidental results directly to physicians. Accomplishments: • Proved it is feasible to establish a structured clinical genetic result delivery network that linking heterogenous organizations in a manner that enabled CDS • Developed e. MERGE XML result transfer format and open sourced via Git. Hub • Published how the network was created to provide a basis for other efforts (Aronson S et al. , JAMIA 2018) • Developed draft FHIR genetic result specification and documentation • Maintained a culture of open collaboration where all sites shared lessons learned relative to ROR and CDS infrastructure Structured results in the EHR or Ancillary OMICs system? 8 7 6 • Site infrastructure presentations 5 • Numerous surveys 3 4 2 • Development and rollout milestone tracking through the project February 2020 Final deliverables 9 1 0 Both Anc EHR None

Milestone 3: EHRI Improve and/or standardize genomic clinical decision support (CDS) for return of clinically relevant genetic or incidental results directly to physicians. Remaining work: • Submit EHRI subgroup manuscripts (June 2020) • NT 213 Rasmussen L et al. Infobutton genomic medicine initiatives survey • NT 319 Watkins N et al. Use of Infobuttons to find answers to clinician’s questions in clinical genomics • Submit case series and editorial on genomic medicine implementation manuscript (June 2020) • NT 352 – Multiple Submissions - Case series and editorial on genomic medicine implementation • Two pilot implementations of e. MERGE FHIR genetics spec (June 2020) February 2020 Final deliverables 10

Goals Milestone 4: Phenotyping 1. Manage 25 e III Phenotype Development & Implementation 2. Leverage NLP to improve phenotyping for five conditions Accomplishments to date • e III phenotypes • 25 (100%) developed; 25 (100%) validated • 23 (92%) Completed; 20 released to the public • 2 shared with the Ao. U project: breast cancer & T 2 DM • Natural Language Processing • Long QT- Arrhythmias released to the network • DQC Guideline (to avoid inconsistent/incomplete data) Data Standardization Efforts • OMOP CDM for both structured and unstructured clinical data • Phecode 1. 2 beta released and paper accepted by JMIR • Online community Phe. WAS Dataset Publication • • • 25 11 25 25 23 20 20 15 10 5 0 Algorithms Developed Validated Implemented Completed Released for Public 5 NLP algorithms: as of 02/20 5 5 4 3 2 Nine (9/20) included in the JBI special issue on deep phenotyping Mapping ICD-10 and ICD-10 -CM Codes to Phecodes: Workflow Development and 1 Initial Evaluation, Wu P, et. al. , JMIR Med Inform 2019; 7(4): e 14325 0 Association of Genetic Risk of Obesity with Postoperative Complications Using Mendelian Randomization, JR Robinson, et. al. , World journal of surgery 44 (1), 84 -94 February 2020 Final deliverables 25 Total e III phenotypes: As of 02/20 25 Algorithms Developed 1 1 Validated Implemented Completed

Milestone 4: Phenotyping Leveraging Natural Language Processing to Improve Phenotyping Remaining Work Non-NLP • Implement the remaining two algorithms by 02/29 NLP • Collect and share preparatory guidelines for NLP implementation by 02/29 • Release NLP phenotypes (i. e. , CRS, ACO, FH, Lupus) by 03/31 • Implement five NLP phenotypes by 05/31 • Submit NLP-based e. III lessons learned manuscripts by 06/30 February 2020 Final deliverables 12

Milestone 5: Ro. R Explore the challenges involved in identifying at-risk family members and informing them of their potential risk. Accomplishments: • Participant surveys (Participant survey subgroup of Ro. R): Familial sharing questions • NT 349: Family communication following return of positive results (Wynn & Milo Rasouly, data analysis in progress) • HCP R 01 • Questions on familial implications on surveys (data collection almost complete) and interviews (in progress) February 2020 Final deliverables 13

Milestone 5: Ro. R Explore the challenges involved in identifying at-risk family members and informing them of their potential risk. Remaining work: • Participant surveys (Participant survey subgroup of Ro. R): • NT 349: Family communication following return of positive results – finish data analysis and manuscript. (6/1/2020) • HCP R 01 (ends 4/1/2021) • • Complete survey data collection (3/1/2020) Complete conducting interviews (6/1/20200) Concept sheets development (9/1/2020) Manuscripts submitted (3/1/2021) February 2020 Final deliverables 14

Milestone 6: Ro. R & Outcomes Estimate the institutional impact of Ro. R. Accomplishments: • Ro. R • Study on the impact on IRBs (completed): Ethical Considerations Related to Return of Results from Genomic Medicine Projects: The e. MERGE Network (Phase III) Experience. Fossey et. al. , J Pers Med. 2018 • NT 273: Returning genomic results to e. MERGE participants: The who, what, where, and how of disclosure (Weisner, submitted for publication) • NT 322: The Reckoning: What We Found After Return of Results for 25, 000 e. MERGE 3 participants (Leppig, data collected, writing in progress) • NT 277: Operationalizing participant choices about genomic results: Beyond all or none ACMG recommended genes (Hoell & Prows, writing in progress) • NT 300: Understanding the return of results process: Content review of patient summary letters (Lynch & Williams J , writing in progress) • NT 323: Challenges in Returning Results in the e. MERGE consortium (Halverson and Clayton, writing in progress) • NT 332: Network-wide lessons learned from the reporting of negative test results (Sharp & Smith, data collection in progress) • NT 340: Practical and Logistical Challenges Related to Return of Genomic Results in Phase III of the e. MERGE Network – Lessons Learned (Kochan & Kullo, data collection in progress) February 2020 Final deliverables 15

Milestone 6: Ro. R & Outcomes Estimate the institutional impact of Ro. R. Remaining work: • Ro. R: Complete manuscript and submit • NT 322: The Reckoning: What We Found After Return of Results for 25, 000 e. MERGE 3 participants – finish data analysis and manuscript. (5/1/2020) • NT 277: Operationalizing participant choices about genomic results: Beyond all or none ACMG recommended genes – finish manuscript. (3/2/2020) • NT 300: Understanding the return of results process: Content review of patient summary letters – finish manuscript. (4/15/2020) • NT 323: Challenges in Returning Results in the e. MERGE consortium – finish manuscript. (5/1/2020) • NT 332: Network-wide lessons learned from the reporting of negative test results – finish data analysis and manuscript. (6/1/2020) • NT 340: Practical and Logistical Challenges Related to Return of Genomic Results in Phase III of the e. MERGE Network – Lessons Learned – finish data analysis and manuscript. (6/1/2020) • Outcomes: Comprehensively examine test ordering after Ro. R to assess economic impact (June 2020) • NT 296 Peterson et al. Collection and analysis of large-scale outcome measures following targeted next generation sequencing February 2020 Final deliverables 16

Milestone 7: Accomplishments Disseminate lessons learned on the various aspects of genomic medicine implementation. Accomplishments: Published Lessons Learned Manuscripts • • • Lessons Learned: 10 published, 11 in development MCS Harmonizing clinical sequencing and interpretation for the e. MERGE-3 Network (The e. MERGE Consortium, AJHG, 2019) Parental attitudes toward consent and data sharing in biobanks: a multi-site experimental survey (Antommaria et al. , AJOB Empir Bioeth. , 2018) Ethical considerations related to return of results from genomic medicine projects: the e. MERGE Network (Phase III) experience (Fossey et al. , J Pers Med, 2018) Harmonizing outcomes for genomic medicine: comparison of e. MERGE outcomes to Clin. Gen outcome/intervention pairs (Williams et al. , Healthcare (Basel), 2018) Healthcare provider education to support integration of pharmacogenomics in practice: the e. MERGE Network experience (Rohrer Vitek et al. , Pharmacogenomics, 2017). Pharmacogenomic clinical decision support design and multi-site process outcomes analysis in the e. MERGE Network (Herr et al. , J Am Med Inform Assoc. , 2019). OMOP Information Model for Phenotyping (Hripcsak et al. , J Biomed Inform. , 2019)* Detecting time-evolving phenotype topics via tensor-factorization on EHRs: Cardiovascular disease case study (Zhao et al. , J Biomed Informatics, 2019; site specific publication)* Empowering genomic medicine by establishing critical sequencing result data flows: the e. MERGE example (Aronson et al. , J Am Med Inform Assoc, 2018). Genomic information for clinicians in the EHR: Lessons learned from Clin. Gen and e. MERGE (Williams et al. Frontiers in Genetics. 2019) February 2020 Final deliverables 10 8 Number publications • 6 4 2 0 Clinical Annotation EHRI Genomics Published Outcomes PGx Phenotyping ROR In development *A subset of the lessons learned publications are also being included in a special issue on Phenotyping Methodology in the Journal of Biomedical Informatics. 17

Milestone 7: Remaining work Disseminate lessons learned on the various aspects of genomic medicine implementation (Network & CC). Remaining work: • Publish NT 357 “Lessons from the e. MERGE Network: Balancing genomics in discovery and in practice. ” Targeted submission early March 2020. • Update website with final lessons learn summaries and publications. Targeted update March 2020. • Support publication process of workgroup lessons learned papers. Full support through June 2020, ongoing tracking of remaining lessons learned publications. February 2020 Final deliverables 18

Other accomplishments: PGx Accomplishments: • Tracked PGx return during e 3 • Established relationship with CPIC, shared lessons learned at in-person meeting and learned what they most wanted from this consortium (to generate additional association information to be used in guideline statements) • Explored potential collaborations with IGNITE • Examined feasibility of additional e 3 PGx outcomes or SPHINX PGx projects • Contributed lessons learned in concert with other working groups (ROR, EHRI) • Publications February 2020 Final deliverables 19

Other remaining work: PGx Remaining work: • Publications in process • Star allele PGx nomenclature (NT 335) • Provide guidance for any PGx work in e 4 • Early prioritization of complicated PGx phenotypes (with subphenotypes for patient and practioner CDS interaction, if applicable) • Consider limitations of current guidelines in non-white populations • Be cautious about assuming any “easy” reuse of existing CDS February 2020 Final deliverables 20

Other accomplishments: Genomics Accomplishments: • Manuscript detailing the genotyping and imputation of ~84, 000 e. MERGE subjects has been published (Stanaway et al; Genetic Epidemiology, 2019). • Provided guidance to the e. MERGE CC regarding genetic data activities in order to produce four large multiple use discovery-based datasets • Co-chair Sleinman led supplement-funded efforts to link Geocoding data to e. MERGE participants. • Establish focus groups for SPHINX Phe. WAS data integration, used focus group data to generate and integrate pilot Phe. WAS data • Assist VUMC with implementation of alternate methods of assessing penetrance and risk-scores using network wide data • Survey sites for availability and release of e. MERGE II/III array data for SV/CNV calling February 2020 Final deliverables 21

Other remaining work: Genomics Accomplishments: • Conduct SV/CNV calling for available array data and determine funding needs for future data analysis • Conducted on Phase I data • Examining funding opportunities for data transfer and analysis of Phase II and III array data • Conduct Phe. WAS on Structural Variant/Copy Number Variant datasets • MCS NT 314 Stanaway et al. The Development of an Imputed Structural Variant Genomic Dataset and Association to Neurological and Alcohol Use Disorder Electronic Medical Record Phenotypes with Biobank Scale Subject Ascertainment • Completed, manuscript in preparation. February 2020 Final deliverables 22

Pediatric accomplishments Describe the opportunities and challenges for pediatric sites in e. MERGE III Accomplishments: • Collection of outcomes data at pediatric sites – similar issues to adult sites • Collection of participant survey data - parents, adolescent participants (13 y – 17 y) and participants 18 y and older • Cascade screening – surveys (CCHMC) and interviews (CHOP) of parents regarding testing of parents and siblings • Approaches established at pediatric sites for return of adult-onset only conditions February 2020 Final deliverables 23

Pediatric other remaining work Describe the opportunities and challenges for pediatric sites in e. MERGE III Remaining work: • Develop manuscript on pediatric (CCHMC, CHOP) results – small N so present cases • Includes cascade testing – robust data collected regarding testing of parents and siblings • Includes discussion of data custody when participant turns 18 years – particular issue for reclassified variants • Includes outcomes data • Address challenges integrating data from pediatric sites with adult sites: Reconciling participant survey data since person completing the survey is the parent for the most part and not the participant • NT 300 – approach to adult-onset only conditions at the pediatric sites (CCHMC, CHOP) and in the Baby. Seq project (BCH) February 2020 Final deliverables 24

June Webinar & Closeout • Final call in June to review progress on the remaining work • June 4, 2020 from 2: 00 -5: 00 pm EST • Generalized Network support (workgroup calls, coordination) will continue through June of 2020 • CC will support individual needs on a case by case basis after that time • Investigators should continue to update publication tracker February 2020 Final deliverables 25

Network-wide milestones 1. Expand the understanding of penetrance by describing the lessons learned from e. MERGE, for example the sample size and age at onset required for penetrance analysis. Conduct penetrance analysis in conditions with sufficient data in the e. MERGE cohort to assess impact on clinical outcomes 2. Determine the impact of return of genetic results (Ro. R) on patients’ immediate outcomes, 6 months and when available 12 months after Ro. R for variants with sufficient prevalence and data, which includes identifying Modification of clinical care (such as changes in prescriptions, lab tests ordered, etc. ) and Outcomes related to processes of care, clinical utility, family utility, provider utility, and patients’ psychosocial factors. 3. Improve and/or standardize genomic clinical decision support (CDS) for return of clinically relevant genetic or incidental results directly to physicians, including initial assessment of impact on relevant process outcomes. 4. Develop a natural language processing (NLP) component for a maximum of five high-priority phenotypes, agreed upon by the phenotyping WG, the Steering Committee and NHGRI. 5. Explore the challenges involved in identifying at-risk family members and informing them of their potential risk as well as collect the responses of the family members. 6. Estimate the institutional impact of Ro. R 7. Disseminate lessons learned on the various aspects of genomic medicine implementation by activities such as publishing articles that propose the key elements for effectively returning genomic results to providers and patients and comparing the impact different methods of Ro. R have on patient and physician care across all sites. February 2020 Final deliverables 26