Nuclear Organization and Gene Expression Section, Laboratory of Biochemistry and Genetics, National Institute of Diabetes and Digestive and Kidney Diseases (NIDDK), National Institutes of Health (NIH), Bethesda, MD, USA.
Bioinformatics. 2022 Sep 2;38(17):4062-4069. doi: 10.1093/bioinformatics/btac498.
ChIP-seq detects protein-DNA interactions within chromatin, such as that of chromatin structural components and transcription machinery. ChIP-seq profiles are often noisy and variable across replicates, posing a challenge to the development of effective algorithms to accurately detect differential peaks. Methods have recently been designed for this purpose but sometimes yield conflicting results that are inconsistent with the underlying biology. Most existing algorithms perform well on limited datasets. To improve differential analysis of ChIP-seq, we present a novel Differential analysis method for ChIP-seq based on Limma (DiffChIPL).
DiffChIPL is adaptive to asymmetrical or symmetrical data and can accurately report global differences. We used simulated and real datasets for transcription factors (TFs) and histone modification marks to validate and benchmark our algorithm. DiffChIPL shows superior performance in sensitivity and false positive rate in different simulations and control datasets. DiffChIPL also performs well on real ChIP-seq, CUT&RUN, CUT&Tag and ATAC-seq datasets. DiffChIPL is an accurate and robust method, exhibiting better performance in differential analysis for a variety of applications including TF binding, histone modifications and chromatin accessibility.
https://github.com/yancychy/DiffChIPL.
Supplementary data are available at Bioinformatics online.
ChIP-seq 可检测染色质内的蛋白质-DNA 相互作用，如染色质结构成分和转录机制。ChIP-seq 图谱在复制过程中通常存在噪声和可变性，这给开发有效算法以准确检测差异峰带来了挑战。为此，最近设计了一些方法，但有时会产生与基础生物学不一致的相互矛盾的结果。大多数现有的算法在有限的数据集上表现良好。为了改善 ChIP-seq 的差异分析，我们提出了一种基于 Limma 的新型 ChIP-seq 差异分析方法（DiffChIPL）。
DiffChIPL 适用于不对称或对称数据，可以准确报告全局差异。我们使用模拟数据集和转录因子 (TF) 和组蛋白修饰标记的真实数据集来验证和基准测试我们的算法。DiffChIPL 在不同的模拟和对照数据集的灵敏度和假阳性率方面表现出优越的性能。DiffChIPL 还可在真实的 ChIP-seq、CUT&RUN、CUT&Tag 和 ATAC-seq 数据集上良好运行。DiffChIPL 是一种准确而稳健的方法，在 TF 结合、组蛋白修饰和染色质可及性等多种应用的差异分析中表现出更好的性能。
https://github.com/yancychy/DiffChIPL。
补充数据可在 Bioinformatics 在线获取。
