Integrating multidimensional genomic data to discover clinically-relevant predictive models

整合多维基因组数据以发现临床相关的预测模型

• 批准号: 9901758
• 负责人: Brittany Nicole Lasseigne
• 金额: $ 24.9万
• 依托单位: UNIVERSITY OF ALABAMA AT BIRMINGHAM
• 依托单位国家: 美国
• 财政年份: 2019
• 资助国家: 美国
• 起止时间: 2019-06-01 至 2022-03-31
• 项目状态: 已结题
• 来源: https://reporter.nih.gov/project-details/9901758
• 关键词: Academic Medical Centers; Advisory Committees; Award; Bioconductor; Biological; Biological Assay; Biotechnology; Budgets; Cancer Etiology; Cell Proliferation; Cells; Characteristics; Chemotherapy-Oncologic Procedure; Chromosomal Instability; Chromosomes; Classification; Clinical; Clinical Data; Collaborations; Complex; Copy Number Polymorphism; Coupling; CpG Island Methylator Phenotype; DNA; DNA Methylation; Data; Data Analyses; Data Set; Disease; Educational process of instructing; Environment; Faculty; Future; Gene Expression; Gene Proteins; Genes; Genetic; Genomic Instability; Genomics; Goals; Heterogeneity; Imagery; Individual; Institutes; Instruction; Laboratories; Lasso; Link; Machine Learning; Malignant Neoplasms; Measurement; Measures; Mental disorders; Mentors; Methodology; Methylation; MicroRNAs; Microsatellite Instability; Modeling; Molecular Profiling; Neurons; Outcome; Pathway interactions; Performance; Phase; Phenotype; Procedures; Regulatory Element; Reproducibility; Research; Research Design; Research Personnel; Resources; Running; Sample Size; Sampling; Scientist; Sensitivity and Specificity; Signal Transduction; Standardization; Structure; Systems Biology; Techniques; Testing; Training; Training Programs; Translational Research; United States National Institutes of Health; Universities; Validation; Variant; Work; Yang; biomarker performance; career development; clinically relevant; computer science; computerized tools; data integration; data reduction; experimental study; genomic data; genomic signature; human disease; improved; insight; multidimensional data; multiple omics; novel strategies; predictive modeling; programs; promoter; protein metabolite; response; single cell sequencing; statistics; tenure track; tool

The goal of this NIH Pathway to Independence award is to provide Dr. Brittany Lasseigne with an extensive training program to prepare her to be an effective independent investigator who uses computational genomics to study complex human diseases. We propose a formal one-year training and mentoring program in genomics, computer science, statistics, and career development to build on her 8+ years of hands-on training, followed by a three-year structured and independent research program. Research will focus on the integration of multidimensional genomic data sets in the context of complex human diseases. A critical barrier in genomic research is the complexity of data integration: the ability to leverage overlapping and unique information captured by different genomic assays would improve our understanding of data integration and generate clinically relevant genomic signatures. To meet this need, we propose to integrate a combination of genomic data we generated with public data to (1) infer genomic instability signatures from different data types, (2) improve clinically relevant phenotype prediction by building multi-omics machine learning classifiers and reducing phenotype heterogeneity, and (3) create a cloud-enabled R package and associated Shiny application to accelerate future research. The proposed work will advance our understanding of data integration, allow inference of genomic instabilities across data sets, and generate high performance classifiers for assessing clinically relevant phenotypes in both cancer and psychiatric disease using frameworks that will be broadly applicable across other complex diseases. It will also facilitate prioritization of experiments in future studies by informing on the orthogonality of genomic assays, thereby allowing more efficient study designs to capture as much information as possible within a given sample size or scope of experimentation. Collectively, this additional training will allow Dr. Lasseigne to develop new multidimensional data integration approaches and translational questions applicable across complex diseases when independent. Dr. Richard Myers (HudsonAlpha) and Dr. Gregory Cooper (HudsonAlpha), leaders in applying genetics and genomics to complex human diseases, and an Advisory Committee of additional experts including Dr. Barbara Wold (Caltech), Dr. Eddy Yang (UAB), and Dr. Timothy Reddy (Duke), will provide mentoring throughout this award. The mentored phase will take place at the HudsonAlpha Institute for Biotechnology, an ideal environment for this training with extensive translational science collaborations, expert faculty and staff, and state-of-the art computational and laboratory resources devoted to genomics. This combination will maximize Dr. Lasseigne's training program, facilitating her transition to an independent, tenure-track investigator at a university with a strong commitment to data-driven approaches to complex human disease research, i.e. strong genomics research programs with clinical collaborators, ideally at, or affiliated with, an academic medical center.

NIH 独立之路奖的目标是为 Brittany Lasseigne 博士提供广泛的 培训计划，让她成为一名使用计算基因组学的有效的独立研究者 研究复杂的人类疾病。我们提出了为期一年的正式培训和指导计划 基因组学、计算机科学、统计学和职业发展以她 8 年多的实践培训为基础， 随后是为期三年的结构化独立研究计划。研究将集中于整合 复杂人类疾病背景下的多维基因组数据集。基因组学的关键障碍 研究的是数据集成的复杂性：利用重叠和独特信息的能力 通过不同的基因组分析捕获将提高我们对数据整合的理解并生成 临床相关的基因组特征。为了满足这一需求，我们建议整合基因组学的组合 我们使用公共数据生成的数据，用于 (1) 从不同数据类型推断基因组不稳定性特征，(2) 通过构建多组学机器学习分类器来改善临床相关表型预测 减少表型异质性，(3) 创建支持云的 R 包和相关的 Shiny 应用以加速未来的研究。拟议的工作将增进我们对数据的理解 集成，允许推断跨数据集的基因组不稳定性，并生成高性能分类器 使用框架评估癌症和精神疾病的临床相关表型 广泛适用于其他复杂疾病。它还将有助于确定未来实验的优先顺序 通过告知基因组测定的正交性来进行研究，从而允许更有效的研究设计 在给定的样本量或实验范围内捕获尽可能多的信息。总的来说， 这项额外的培训将使 Lasseigne 博士能够开发新的多维数据集成方法 以及独立时适用于复杂疾病的转化问题。理查德·迈尔斯博士 (HudsonAlpha) 和 Gregory Cooper 博士 (HudsonAlpha)，是将遗传学和基因组学应用于 复杂的人类疾病，以及包括 Barbara Wold 博士在内的其他专家组成的咨询委员会 （加州理工学院）、Eddy Yang 博士（阿拉巴马大学）和 Timothy Reddy 博士（杜克大学）将在整个颁奖过程中提供指导。 指导阶段将在 HudsonAlpha 生物技术研究所进行，这是一个理想的环境 该培训涉及广泛的转化科学合作、专家教职人员和最先进的技术 专门用于基因组学的计算和实验室资源。这种组合将最大限度地提高 Lasseigne 博士的 培训计划，帮助她过渡到一所大学的独立终身教授研究员 坚定致力于以数据驱动的方法进行复杂的人类疾病研究，即强大的基因组学 与临床合作者合作的研究项目，最好是在学术医学中心或附属于学术医学中心。