Fig. 5
From: Identifying peaks in *-seq data using shape information
Cross-correlation plots of two low-quality ChIP-seq datasets. On the left, the read-length peak is significantly higher than the fragment-length peak (relative strand correlation of around 0.5), indicating potential problems in the immune-precipitation step. On the right, a very noisy cross-correlation profile indicates a ChIP-seq experiment where a very small number of reads was sequenced