MAnorm2 for quantitatively comparing groups of ChIP-seq samples

Abstract

Eukaryotic gene transcription is regulated by a large cohort of chromatin associated proteins, and inferring their differential binding sites between cellular contexts requires a rigorous comparison of the corresponding ChIP-seq data. We present MAnorm2, a new computational tool for quantitatively comparing groups of ChIP-seq samples. MAnorm2 uses a hierarchical strategy to normalize ChIP-seq data and then performs differential analysis by assessing within-group variability of ChIP-seq signals under an empirical Bayes framework. In this framework, MAnorm2 considers the abundance of differential ChIP-seq signals between groups of samples and the possibility of different within-group variability between groups. When samples in each group are biological replicates, MAnorm2 can reliably identify differential binding events even between highly similar.

A systematic dissection of the epigenomic heterogeneity of lung adenocarcinoma reveals two different subclasses with distinct prognosis and core regulatory networks

“…We next processed the ChIP-seq data for each sample as previously described before further analysis [ 28 , 29 ]. Briefly, after data quality control, reads alignment, peak calling, and filtering, in total we identified ~ 100 thousand H3K27ac peaks for LUAD samples and ~ 60 thousand peaks for normal samples, mostly located at intergenic and intronic regions (Additional file 1 : Figures S1A and S1C).…”

“…1 a, Additional file 1 : Figure S1B). We grouped the H3K27ac ChIP-seq profiles of LUAD and normal lung samples, and then used the MAnorm2 model to compare these two groups of profiles [ 29 ]. By this method, we detected a huge number of differential H3K27ac sites, including 4784 tumor-specific and 7645 normal-specific ones (Fig.…”

“…We next quantitatively evaluated the intertumoral variations of H3K27ac signals normalized by MAnorm2 on genome scale. In order to remove the mean-variance dependence, we fitted a mean-variance curve and selected those peaks with significant deviations above the curve [ 29 ]. By this method, we identified 4615 hyper-variable H3K27ac peaks (HVPs) among the tumor samples (Fig.…”

“…H3K27ac peaks were called by using MACS v1.3.7, and the 30,000 most significant peaks associated with each sample were retained. All H3K27ac peaks were merged into a consensus list of peaks, and reads finally within these peaks were counted by using MAnorm2-utils [ 29 ]. Differential H3K27ac sites between tumor and normal samples (or between GII and GI) were identified using MAnorm2 [ 29 ].…”

“…All H3K27ac peaks were merged into a consensus list of peaks, and reads finally within these peaks were counted by using MAnorm2-utils [ 29 ]. Differential H3K27ac sites between tumor and normal samples (or between GII and GI) were identified using MAnorm2 [ 29 ]. Those sites with adjusted p -value below 0.05 and fold change greater than 2 were defined as differential ones.…”