权益分类	功能权益	普通用户	{{item.name}}会员
{{category.name}}	{{benefitItem.name}}

Integration of Clinical and Omic Data for Improved Prediction of Patient Outcome

整合临床和组学数据以改进对患者结果的预测

基本信息

批准号：
9469716
负责人：
Matthew Ruffalo
金额：
$ 5.9万
依托单位：
CARNEGIE-MELLON UNIVERSITY
依托单位国家：
美国
项目类别：
财政年份：
2018
资助国家：
美国
起止时间：
2018-02-05 至 2021-02-04
项目状态：
已结题

来源：
https://reporter.nih.gov/project-details/9469716
关键词：
Algorithms Antineoplastic Agents Big Data Biological Biomedical Technology Breast Cancer Treatment Cancer Patient Categories Cell Line Cell physiology Classification Clinical Clinical Data Collaborations Complex Data Data Analyses Data Set Data Sources Drug Targeting Gene Expression Gene Proteins Genes Genomics Health In Vitro Investigation Knock-out Knowledge Laboratories Machine Learning Malignant Neoplasms Methodology Methods Methylation Modeling Outcome Output Pathogenesis Pathology Report Pathway interactions Patient-Focused Outcomes Patients Performance Pharmaceutical Preparations Process Proteins Proteomics Recommendation Research Personnel Sampling Series Somatic Mutation Symptoms Techniques Time Universities base cancer therapy cancer type cell type clinically relevant data integration differential expression experimental study improved knockout gene learning strategy malignant breast neoplasm medication administration multitask novel predicting response programs response small molecule transcriptomics treatment response

项目摘要

Integration of Clinical and Omic Data for Improved Prediction of Patient Outcome PI: Matthew Ruffalo – Sponsors: Ziv Bar-Joseph, Stefﬁ Oesterreich A Project Summary This project comprises a data integration and machine learning methodology to improve performance in predicting patient outcome – speciﬁcally, response to cancer treatments. Multiple disparate data types will be combined for this task, including omic data (somatic mutations, gene expression, methylation), interaction networks, drug target information, and previously-unavailable clinical data, to produce better predictions of response to speciﬁc cancer treatments. The proposed methods will use relationships between genes and proteins (often represented as protein in- teraction networks) to construct composite features from omic data, and use these features as inputs to machine learning algorithms, to improve prediction performance of classiﬁcation methods when applied to clinically rel- evant prediction tasks. Additionally, this project will use clinical data at a level that is not typically available for breast cancer samples, obtained via the Center for Big Data for Better Health (BD4BH) collaboration between Carnegie Mellon University and the University of Pittsburgh. This data includes time-series clinical data spanning ﬁve years of breast cancer treatment, such as medication administration, laboratory results, pathology reports, symptoms, and other types of data. This clinical data will be integrated into composite features in order to fur- ther improve prediction performance. This feature construction methodology will also allow the investigation of which cellular processes and pathways are most strongly associated with clinical outcomes, such as response to speciﬁc treatments and patient survival. While this approach shows promise in improving classiﬁcation/prediction performance in clinically relevant tasks such as survival and response to treatment, models learned from these integrated features may still overﬁt the classiﬁcation task and may not generalize to other cancers or drugs with similar mechanisms of action. As such, this method will integrate cell line expression data from the LINCS project into the predictive features. The LINCS program has proﬁled gene expression changes in cell lines under two broad categories: introduction of small molecules, and gene knockouts. Both sets of data will be used to constrain the sets of genes that are used as features for prediction of response to treatment of certain drugs. A central hypothesis of this proposal is that many cellular processes and cancer types respond in similar ways to such perturbations, allowing the use of a multi-task learning method to identify the commonalities in cellular response to these drugs across cell lines. Such methods will also allow for identiﬁcation of cell-type-speciﬁc and cancer-speciﬁc responses to such perturbations, identifying those networks and processes that speciﬁcally relate to response to treatment in speciﬁc cancers. Results will be validated via new in vitro cell line experiments, demonstrating that the constructed features are informative in clinical settings. 1

整合临床和Omic数据以改善患者结局的预测 PI：Matthew Ruffalo -赞助商：Ziv Bar-Joseph，Stef FiOesterreich 项目摘要该项目包括数据集成和机器学习方法，以提高预测性能患者结局-特别是对癌症治疗的反应。将组合多种不同的数据类型，这项任务，包括组学数据（体细胞突变，基因表达，甲基化），相互作用网络，药物靶点信息和以前无法获得的临床数据，以更好地预测对特定癌症的反应治疗。所提出的方法将使用基因和蛋白质之间的关系（通常表示为蛋白质，交互网络）从组学数据构建复合特征，并将这些特征用作机器的输入。学习算法，以提高分类方法的预测性能，当应用于临床相关预测任务。此外，本项目将使用通常无法获得的临床数据，乳腺癌样本，通过大数据促进健康中心（BD 4 BH）与卡内基梅隆大学和匹兹堡大学。该数据包括时间序列临床数据，五年的乳腺癌治疗，如药物管理，实验室结果，病理报告，症状和其他类型的数据。该临床数据将被整合到复合特征中，以便进一步- 从而提高预测性能。这种特征构造方法还将允许调查哪些细胞过程和途径与临床结果最密切相关，例如对具体治疗和患者生存。虽然这种方法在改善临床相关疾病的分类/预测性能方面显示出希望，例如生存和对治疗的反应等任务，从这些综合特征中学习的模型可能仍然过拟合。分类任务，可能无法推广到其他癌症或具有类似作用机制的药物。作为因此，该方法将来自LINCS项目的细胞系表达数据整合到预测特征中。的 LINCS计划将细胞系中的基因表达变化分为两大类：小分子和基因敲除。这两组数据将用于约束所使用的基因组作为预测对某些药物治疗的反应的特征。这一提议的一个核心假设是，许多细胞过程和癌症类型以类似的方式对这种扰动作出反应，允许使用多任务学习方法，以确定跨细胞系对这些药物的细胞反应的共性。等这些方法还将允许识别对这种扰动的细胞类型特异性和癌症特异性反应，识别那些与特定癌症治疗反应特别相关的网络和过程。结果将通过新的体外细胞系实验进行验证，证明构建的特征是有效的。在临床环境中提供信息。 1