Notes on Yao, Liang, Ozer, Leung, Lis, Yu, A comparison of experimental assays and analytical methods for genome-wide identification of active enhancers

Yu₂₀₂₂

A comparison of experimental assays and analytical methods for genome-wide identification of active enhancers

Introduction

• Distal divergent transcription events are more reliable marks for activate enhancers than epigenomics signatures

  • To identify active enhancers genome wide it's critical to detect eRNAs and their TSSs with high sensitivity

• Conventional RNA-seq experiments capture eRNAs with very low efficiency

• TSS Assays vs Nascent Transcript Assays

Fig. 1 Comparison of currently available assays for detection of eRNAs.

Results

TSS assays are more sensitive in regard to eRNA detection.

Fig. 2 Evaluation of assay sensitivity in eRNA detection.

Gene body reads in NT assays contribute to lower sensitivity.

Cap selection or Pol II pausing does not bias eRNA capture.

Fig. 3 Characterization of factors affecting assay sensitivity and evaluation of assay specificity in eRNA detection.

TSS assays, especially GRO-cap, have higher specificity.

PINTS.

PINTS achieves high overall performance for TSS assays.

Peak callers vary in resolution and computational needs.

An enhancer compendium for human cell and tissue types.

Fig. 4 PINTS achieves optimal balance among resolution, robustness, sensitivity, specificity and computational resources required.

Fig. 5 A comprehensive human enhancer compendium.

PINTS web server for exploration and analysis of enhancers.

Fig. 6 Interactive PINTS web server.

Discussion

Online content

Extended Data Fig. 1 An extended evaluation of eRNA detection sensitivity of different assays.

Extended Data Fig. 2 Effect of technical artifacts on eRNA capture.

Extended Data Fig. 3 Analyses of factors affecting assays’ sensitivity in detecting eRNAs.

Extended Data Fig. 4 Extended evaluations of assays’ specificity.

Extended Data Fig. 5 Assessments of transcript unit prediction and schematic illustration of PINTS.

Extended Data Fig. 6 Profiles of peak calls generated by different peak callers for various assays.

Extended Data Fig. 7 Extended analyses on the robustness of element predictions.

Extended Data Fig. 8 Performance evaluation of peak callers under different sequencing depths.

Extended Data Fig. 9 Profiles of unique distal elements identified by different tools.

Extended Data Fig. 10 A summary of the computational tools compared in this study.