Impulse model-based differential expression analysis of time course sequencing data
Impulse model-based differential expression analysis of time course sequencing data
复制标题
DOI：
10.1101/113548
复制
发表时间：
2017-03
影响因子：
14.9
通讯作者：
David S. Fischer;Fabian J Theis;N. Yosef
中科院分区：
文献类型：
--
作者：
David S. Fischer;Fabian J Theis;N. Yosef
细胞系统响应于发育或环境刺激的全局基因表达轨迹通常遵循可以用脉冲模型建模的原型单脉冲或状态转换模式。在这里，我们联合收割机将连续脉冲表达模型与ImpulseDE 2中的测序数据噪声模型相结合，ImpulseDE 2是一种用于时间过程测序实验（如RNA-seq，ATAC-seq和ChIP-seq）的差异表达算法。我们表明，ImpulseDE 2优于目前使用的差分表达算法的数据集有足够多的采样时间点。ImpulseDE 2能够区分瞬时和单调变化的表达轨迹。这种分类将负责初始和最终细胞状态表型的基因与驱动细胞状态转变或由细胞状态转变驱动的基因分开，并将氧化磷酸化的下调鉴定为可驱动人胚胎干细胞分化的分子标记。
The global gene expression trajectories of cellular systems in response to developmental or environmental stimuli often follow the prototypic single-pulse or state-transition patterns which can be modeled with the impulse model. Here we combine the continuous impulse expression model with a sequencing data noise model in ImpulseDE2, a differential expression algorithm for time course sequencing experiments such as RNA-seq, ATAC-seq and ChIP-seq. We show that ImpulseDE2 outperforms currently used differential expression algorithms on data sets with sufficiently many sampled time points. ImpulseDE2 is capable of differentiating between transiently and monotonously changing expression trajectories. This classification separates genes which are responsible for the initial and final cell state phenotypes from genes which drive or are driven by the cell state transition and identifies down-regulation of oxidative-phosphorylation as a molecular signature which can drive human embryonic stem cell differentiation.