A pan-cancer atlas of driver mutations in >100,000 patients based on a hypothesis-driven combined computational and experimental approach

基于假设驱动的计算和实验相结合的方法，绘制了超过 100,000 名患者的驱动突变泛癌图谱

• 批准号: 10276520
• 负责人: Felix Dietlein
• 金额: $ 13.32万
• 依托单位: DANA-FARBER CANCER INST
• 依托单位国家: 美国
• 财政年份: 2021
• 资助国家: 美国
• 起止时间: 2021-08-16 至 2022-03-31
• 项目状态: 已结题
• 来源: https://reporter.nih.gov/project-details/10276520
• 关键词: Accounting; Address; Advanced Malignant Neoplasm; Affect; Atlases; Binding; Biological; Biology; CRISPR interference; Cancer Patient; Cells; Chromatin Remodeling Factor; Clinical; Clinical Markers; Clustered Regularly Interspaced Short Palindromic Repeats; Code; Complex; Computer software; Computing Methodologies; Data; Data Set; Development; Development Plans; Diagnostic; Drug Targeting; Environment; Estrogens; Event; Foundations; Future; Genes; Genetics and Medicine; Genets; Genome; Genomics; Goals; Immunotherapy; Individual; Institutes; Lead; Leadership; Malignant Neoplasms; Manuals; Maps; Medicine; Mentors; Methods; Mission; Modeling; Mutation; Oncogenic; Open Reading Frames; Outcome; Pathologic Mutagenesis; Pathway interactions; Patients; Proteins; Public Health; Publications; Research; Role; Scientist; Signal Transduction; Solid; Somatic Mutation; Statistical Algorithm; Statistical Methods; Statistical Models; Structure; System; Techniques; Testing; The Cancer Genome Atlas; Training; United States National Institutes of Health; Untranslated RNA; base; cancer gene expression; cancer genome; cancer genomics; cancer survival; cancer therapy; career development; clinical care; clinical predictors; computerized tools; driver mutation; exome; experience; experimental study; genome editing; genome sequencing; genomic biomarker; improved; innovation; interdisciplinary approach; machine learning method; malignant breast neoplasm; mathematical methods; mathematical model; medical schools; meetings; mid-career faculty; new therapeutic target; novel; novel therapeutics; open source; precision oncology; professor; promoter; skills; symposium; targeted treatment; tool; transcription factor; tumor; tumorigenesis; whole genome

PROJECT SUMMARY Most mutations in cancer genomes are random passengers that do not contribute to oncogenesis, whereas only a few are drivers critical for tumor development. Existing cancer therapies interfere directly with the biology of drivers, which have been characterized extensively in protein-coding regions but remain largely uncharacterized outside coding regions. Most tumors harbor a combination of several driver mutations, but it is unclear how multiple events are coordinated in tumor development. The applicant's long-term goal is to advance cancer medicine by identifying new drug targets and clinical markers for therapies in complex pathways. The overall objectives in this application are to (i) reveal the biological role of noncoding drivers, (ii) capture the coordination of driver events at a pathway level, and (iii) profile the effects of noncoding drivers on cancer gene expression. The central hypothesis is that refining the biological assumptions of computational methods will enhance their statistical power. The rationale is that defining the biology of noncoding drivers and their combination will offer a strong foundation for new therapies. The central hypothesis will be tested in three specific aims: 1) Determine the impact of integrating biological mechanisms into statistical methods for localizing noncoding drivers; 2) Evaluate mechanisms by which promoter mutations increase the expression of cancer genes; and 3) Assess the coordination of multiple driver events in tumor development. The proposed research is innovative, in the applicant's opinion, because it will allow for an unbiased characterization of driver mutations across the entire genome, address the limitations of existing cancer genomics methods in noncoding regions, and facilitate the usage of statistical concepts for non-computational scientists. The proposal is significant because it will enable a systematic interrogation of noncoding drivers and their combinations. Ultimately, this will pave the way for new targeted therapies. Dr. Dietlein will be mentored by Dr. Van Allen, an Associate Professor of Medicine at Harvard Medical School with considerable experience in cancer genomics methods that require statistical innovation for clinically focused questions. His co-mentor, Dr. Meyerson, is a Professor of Genetics and Medicine at Harvard Medical School and a pioneer in developing targeted therapies based on driver mutations. Additional support will be provided by 4 computational and 2 experimental collaborators. Dr. Dietlein's training plan contains four goals, which will be pursued by hands-on experiential training, conference meetings, and structured coursework: 1) Acquire computational skills for interpreting drivers in noncoding regions; 2) Experimental techniques to validate driver mutations by CRISPR interference; 3) Develop professional leadership skills for interdisciplinary teams of scientists; and 4) Use machine-learning methods for interpreting drivers in cancer genomes. Dana-Farber, Harvard Medical School, and the Broad Institute provide an ideal environment to execute the applicant's career development plan.

项目概要 癌症基因组中的大多数突变都是随机乘客，不会导致肿瘤发生，而 只有少数是肿瘤发展的关键驱动因素。现有的癌症疗法直接干扰 驱动生物学，其在蛋白质编码区域已被广泛表征，但在很大程度上仍然存在 未表征的外部编码区域。大多数肿瘤都包含几种驱动突变的组合，但它是 尚不清楚肿瘤发展中多个事件如何协调。申请人的长期目标是 通过确定复杂治疗的新药物靶点和临床标志物来推进癌症医学 途径。该应用程序的总体目标是（i）揭示非编码驱动程序的生物学作用，（ii） 捕获路径级别驱动事件的协调，以及 (iii) 分析非编码驱动因素对 癌症基因表达。中心假设是完善计算的生物学假设 方法将增强其统计能力。基本原理是定义非编码驱动程序的生物学和 他们的组合将为新疗法奠定坚实的基础。中心假设将在三个方面进行检验 具体目标： 1) 确定将生物机制整合到统计方法中的影响 本地化非编码驱动程序； 2) 评估启动子突变增加表达的机制 癌症基因； 3) 评估肿瘤发展中多个驱动事件的协调性。拟议的 申请人认为，这项研究具有创新性，因为它将能够对驾驶员进行公正的描述 整个基因组的突变，解决了现有非编码癌症基因组学方法的局限性 区域，并促进非计算科学家使用统计概念。该提案是 意义重大，因为它将能够系统地询问非编码驱动程序及其组合。 最终，这将为新的靶向治疗铺平道路。 Dietlein 博士将接受 Van Allen 博士的指导。 哈佛医学院医学副教授，在癌症基因组学方面拥有丰富经验 需要针对临床重点问题进行统计创新的方法。他的共同导师迈耶森博士是一位 哈佛医学院遗传学和医学教授，开发靶向疗法的先驱 基于驱动突变。额外的支持将由 4 个计算和 2 个实验提供 合作者。 Dietlein 博士的培训计划包含四个目标，这些目标将通过实践体验来实现 培训、会议和结构化课程：1) 获得口译计算技能 非编码区域的驱动程序； 2）通过CRISPR干扰验证驱动突变的实验技术； 3）培养跨学科科学家团队的专业领导技能； 4）使用机器学习 解释癌症基因组驱动因素的方法。达纳—法伯癌症研究所、哈佛医学院和博德研究所 学院为申请人执行职业发展计划提供了理想的环境。