NHLBI TOPMed - NHGRI CCDG: The BioMe Biobank at Mount Sinai

The IPM BioMe Biobank, founded in September 2007, is an ongoing, broadly-consented electronic health record (EHR)-linked clinical care biobank that enrolls participants non-selectively from the Mount Sinai Medical Center patient population. BioMe currently comprises >42,000 participants from diverse ancestries, characterized by a broad spectrum of longitudinal biomedical traits. Participants enroll through an opt-in process and consent to be followed throughout their clinical care (past, present, and future) in real-time, allowing us to integrate their genomic information with their EHRs for discovery research and clinical care implementation. BioMe participants consent for recall, based on their genotype and/or phenotype, permitting in-depth follow-up and functional studies for selected participants at any time. Phenotypic and genomic data are stored in a secure database and made available to investigators, contingent on approval by the BioMe Governing Board. BioMe uses a "data-broker" system to protect confidentiality.

Ancestral diversity - BioMe participants represent a broad racial, ethnic and socioeconomic diversity with a distinct and population-specific disease burden. Specifically, BioMe participants are of African (AA), Hispanic/Latino (HL), European (EA) and other/mixed ancestry (Table 1, Figure 1). BioMe participants are predominantly of African (AA, 24%), Hispanic/Latino (HL, 35%), European (EA, 32%), and other ancestry (OA, 10%) (Table 1, Figure 1). Participants who self-identify as Hispanic/Latino further report to be of Puerto Rican (39%), Dominican (23%), Central/South American (17%), Mexican (5%) or other Hispanic (16%) ancestry. More than 40% of European ancestry participants are genetically determined to be of Ashkenazi Jewish ancestry. With this broad ancestral diversity, BioMe is uniquely positioned to examine the impact of demographic and evolutionary forces that have shaped common disease risk.

Phenotypes available in BioMe - BioMe has available a high-quality and validated set of fully implemented clinical phenotype data that has been culled by a multi-disciplinary team of experienced investigators, clinicians, information technologists, data-managers, and programmers who apply advanced medical informatics and data mining tools to extract and harmonize EHRs. BioMe, as a cohort, offers great versatility for designing nested case-control sample-sets, particularly for studying longitudinal traits and co-morbidity in disease burden.

** Biomedical and clinical outcomes: The BioMe Biobank is linked to Mount Sinai's system-wide Epic EHR, which captures a full spectrum of biomedical phenotypes, including clinical outcomes, covariate and exposure data from past, present and future health care encounters. As such, the BioMe Biobank has a longitudinal design as participants consent to make all of their EHR data from past (dating back as far as 2003), present and future inpatient or outpatient encounters available for research, without restriction. The median number of outpatient encounters is 21 per participant, reflecting predominant enrollment of participants with common chronic conditions from primary care facilities.

** Environmental data: The clinical and EHR information is complemented by detailed demographic and lifestyle information, including ancestry, residence history, country of origin, personal and familial medical history, education, socio-economic status, physical activity, smoking, dietary habits, alcohol intake, and body weight history, which is collected in a systematic manner by interview-based questionnaire at time of enrollment.

The IPM BioMe Biobank contributed ~10,600 DNA samples for whole genome sequencing to the TOPMed program. Samples were selected for the Coronary Artery Disease (CAD) and the Chronic Obstructive Pulmonary Disease (COPD) working groups. Using a Case-Definition-Algorithm (CDA), we identified ~4,100 individuals with CAD (~50% women) and ~3,000 individuals as controls (65% women). In addition, we identified ~800 individual with COPD (62% women) and 1800 as controls (72% women). Another 600 BioMe participants with Atrial Fibrillation, all of African ancestry, were included.

IPM BioMe生物样本库（IPM BioMe Biobank）于2007年9月成立，是一项持续开展、获得广泛知情同意的、与电子健康档案（Electronic Health Record, EHR）关联的临床诊疗生物样本库，从西奈山医疗中心（Mount Sinai Medical Center）的患者群体中无选择性招募参与者。目前BioMe拥有超过42000名来自不同祖先群体的参与者，其队列涵盖广泛的纵向生物医学特征。参与者通过知情选择加入流程注册，并同意在其整个临床诊疗过程中（既往、当前及未来）被实时随访，这使得研究团队能够将其基因组信息与电子健康档案数据整合，用于发现性研究与临床诊疗落地实践。BioMe参与者同意基于其基因型和/或表型进行召回，以便在任意时间为选定参与者开展深度随访与功能研究。表型与基因组数据存储于安全数据库中，仅在获得BioMe管理委员会批准后，方可向研究者开放。BioMe采用“数据中介”（data-broker）系统以保护参与者隐私。

祖先多样性——BioMe参与者涵盖广泛的种族、族裔与社会经济多样性，且具有独特的人群特异性疾病负担。具体而言，BioMe参与者分为非洲裔（AA, African American）、西班牙裔/拉丁裔（HL, Hispanic/Latino）、欧洲裔（EA, European Ancestry）及其他/混合血统群体（表1、图1）。具体构成比例为：非洲裔（AA，24%）、西班牙裔/拉丁裔（HL，35%）、欧洲裔（EA，32%）及其他血统群体（OA，10%）（表1、图1）。自我认定为西班牙裔/拉丁裔的参与者中，进一步报告其血统为波多黎各裔（39%）、多米尼加裔（23%）、中美洲/南美洲裔（17%）、墨西哥裔（5%）或其他西班牙裔群体（16%）。超过40%的欧洲裔参与者经基因检测确定为德系犹太血统。凭借如此广泛的祖先多样性，BioMe具备独特优势，可用于探究塑造常见疾病风险的人口统计学与进化因素的影响。

BioMe可获取的表型数据——BioMe拥有一套高质量且经过全面验证的正式上线临床表型数据集，该数据集由多学科团队（包括经验丰富的研究者、临床医师、信息科技专家、数据管理人员与程序员）整合而成，他们运用先进的医学信息学与数据挖掘工具提取并统一标准化电子健康档案数据。作为一个队列研究，BioMe在设计嵌套病例对照样本集方面具有极高的灵活性，尤其适用于研究疾病负担中的纵向特征与共病情况。

** 生物医学与临床结局 **：BioMe生物样本库与西奈山医疗系统的Epic电子健康档案系统相连，该系统可覆盖全范围的生物医学表型，包括来自既往、当前及未来医疗就诊的临床结局、协变量与暴露数据。因此，BioMe生物样本库采用纵向研究设计：参与者同意将其所有电子健康档案数据（最早可追溯至2003年），包括既往、当前及未来的住院或门诊就诊数据，无限制地开放用于研究。参与者的门诊就诊次数中位数为21次，这反映出招募的参与者多为来自基层医疗机构的常见慢性病患者。

** 环境暴露数据 **：临床与电子健康档案信息辅以详细的人口统计学与生活方式信息，包括血统、居住史、原籍国、个人与家族病史、教育程度、社会经济地位、体力活动、吸烟情况、饮食习惯、酒精摄入与体重变化史。这些信息均在参与者注册时通过访谈式问卷系统收集。

IPM BioMe生物样本库为精准医学跨组学（Trans-Omics for Precision Medicine, TOPMed）项目提供了约10600份用于全基因组测序的DNA样本。样本被选取用于冠状动脉疾病（Coronary Artery Disease, CAD）与慢性阻塞性肺疾病（Chronic Obstructive Pulmonary Disease, COPD）工作组。通过病例定义算法（Case-Definition-Algorithm, CDA），研究团队共识别出约4100名冠状动脉疾病患者（其中约50%为女性）与约3000名对照个体（其中65%为女性）。此外，团队还识别出约800名慢性阻塞性肺疾病患者（62%为女性）与1800名对照个体（72%为女性）。另有600名均为非洲裔血统的心房颤动BioMe参与者被纳入该项目。