In this paper, we propose a statistical framework MLDA, based on the smoothed LDA to detect the differential usage of transcript isoforms for RNA-seq data. The experimental results show that our method performs competitively on the detection of DTU and obtain more accurate relative transcript abundance compared with other alternatives on both simulated and real data.

中文翻译：

---

基于平滑的LDA模型检测跨多个条件的RNA-seq数据的差异转录使用

在本文中，我们提出了一种基于平滑LDA的统计框架MLDA，以检测RNA-seq数据的转录异构体的差异用法。实验结果表明，与模拟和真实数据上的其他替代方法相比，我们的方法在DTU的检测上具有竞争优势，并且可以获得相对准确的相对转录本丰度。