Diabetes and Complications Research, Eli Lilly and Company, Indianapolis, USA.
Curr Opin Cardiol. 2019 May;34(3):246-253. doi: 10.1097/HCO.0000000000000611.
Comprehensive analyses of the genome, transcriptome, proteome and metabolome are instrumental in identifying biomarkers of disease, to gain insight into mechanisms underlying the development of cardiovascular disease, and show promise for better stratifying patients according to disease subtypes. This review highlights recent 'omics' studies, including integration of multiple 'omics' that have advanced mechanistic understanding and diagnosis in humans and animal models.
Transcriptome-based discovery continues to be a primary method to obtain data for hypothesis generation and the understanding of disease pathogenesis has been enhanced by single cell-based methods capable of revealing heterogeneity in cellular responses. Advances in proteome coverage and quantitation of individual protein species, together with enhanced methods for detecting posttranslational modifications, have improved discovery of protein-based biomarkers.
High-throughput assays capable of quantitating the vast majority of any particular type of biomolecule within a tissue sample, isolated cells or plasma are now available. In order to make best use of the large amount of data that can be generated on given molecule types, as well as their interrelationships in disease, continued development of pattern-recognition algorithms ('machine learning') will be required and the subclassification of disease that is made possible by such algorithms will be likely to inform clinical practice, and vice versa.
全面分析基因组、转录组、蛋白质组和代谢组有助于识别疾病的生物标志物，深入了解心血管疾病发展的机制，并有望根据疾病亚型更好地对患者进行分层。本综述强调了最近的“组学”研究，包括整合多种“组学”，这些研究促进了人类和动物模型中机制理解和诊断的进展。
基于转录组的发现仍然是获得数据以进行假设生成的主要方法，单细胞方法能够揭示细胞反应的异质性，从而增强了对疾病发病机制的理解。蛋白质组覆盖范围的提高和单个蛋白质种类的定量，以及检测翻译后修饰的增强方法，都提高了基于蛋白质的生物标志物的发现。
现在有高通量检测方法能够定量检测组织样本、分离细胞或血浆中任何特定类型生物分子的绝大多数。为了充分利用给定分子类型以及它们在疾病中的相互关系上可以生成的大量数据，需要继续开发模式识别算法（“机器学习”），并且此类算法实现的疾病细分可能会反过来影响临床实践。
