European Research Institute for the Biology of Ageing, University of Groningen, University Medical Centre Groningen, Groningen, The Netherlands.
Gene Learning Association, Geneva, Switzerland.
Sci Rep. 2021 May 24;11(1):10793. doi: 10.1038/s41598-021-90192-9.
Finding novel biomarkers for human pathologies and predicting clinical outcomes for patients is challenging. This stems from the heterogeneous response of individuals to disease and is reflected in the inter-individual variability of gene expression responses that obscures differential gene expression analysis. Here, we developed an alternative approach that could be applied to dissect the disease-associated molecular changes. We define gene ensemble noise as a measure that represents a variance for a collection of genes encoding for either members of known biological pathways or subunits of annotated protein complexes and calculated within an individual. The gene ensemble noise allows for the holistic identification and interpretation of gene expression disbalance on the level of gene networks and systems. By comparing gene expression data from COVID-19, H1N1, and sepsis patients we identified common disturbances in a number of pathways and protein complexes relevant to the sepsis pathology. Among others, these include the mitochondrial respiratory chain complex I and peroxisomes. This suggests a Warburg effect and oxidative stress as common hallmarks of the immune host-pathogen response. Finally, we showed that gene ensemble noise could successfully be applied for the prediction of clinical outcome namely, the mortality of patients. Thus, we conclude that gene ensemble noise represents a promising approach for the investigation of molecular mechanisms of pathology through a prism of alterations in the coherent expression of gene circuits.
发现人类病理学的新生物标志物并预测患者的临床结果具有挑战性。这源于个体对疾病的异质反应，反映在基因表达反应的个体间可变性上，从而掩盖了差异基因表达分析。在这里，我们开发了一种可以用于剖析与疾病相关的分子变化的替代方法。我们将基因集合噪声定义为一种衡量标准，用于衡量一个集合中基因的方差，这些基因编码已知生物途径的成员或注释蛋白复合物的亚基，并在个体内进行计算。基因集合噪声允许在基因网络和系统的水平上整体识别和解释基因表达失衡。通过比较 COVID-19、H1N1 和脓毒症患者的基因表达数据，我们确定了与脓毒症病理学相关的许多途径和蛋白复合物中的常见紊乱。其中包括线粒体呼吸链复合物 I 和过氧化物酶体。这表明有氧糖酵解和氧化应激是免疫宿主-病原体反应的共同特征。最后，我们表明基因集合噪声可以成功地应用于预测临床结果，即患者的死亡率。因此，我们得出结论，基因集合噪声代表了一种通过改变基因电路的协调表达来研究病理学分子机制的有前途的方法。
