PATTY corrects open-chromatin bias for improved bulk and single-cell CUT&Tag profiling.

Area of Science:

• Genomics
• Epigenetics
• Computational Biology

Background:

• Accurate epigenomic profiling is crucial for understanding gene regulation.
• Cleavage Under Targets & Tagmentation (CUT&Tag) is a widely used method for epigenomic profiling, especially for low-input and single-cell samples.
• CUT&Tag is susceptible to open-chromatin bias due to Tn5 transposase, which can distort genome-wide signals and complicate downstream analysis, particularly in sparse single-cell data.

Purpose of the Study:

• To address the open-chromatin bias inherent in CUT&Tag experiments.
• To develop a computational method for correcting this bias.
• To improve the accuracy and robustness of epigenomic signal detection and downstream analysis, especially for single-cell applications.

Main Methods:

• We developed PATTY (Propensity Analyzer for Tn5 Transposase Yielded bias), a machine learning-based computational method.
• PATTY leverages accompanying ATAC-seq data to correct CUT&Tag open-chromatin bias.
• The method was validated on published CUT&Tag datasets and applied to single-cell CUT&Tag analyses.

Main Results:

• Open-chromatin bias was confirmed to be prevalent in published CUT&Tag data, even with optimized protocols.
• PATTY successfully corrects open-chromatin bias, enabling accurate signal detection for active and repressive histone modifications (H3K27ac, H3K27me3, H3K9me3).
• In single-cell CUT&Tag data, PATTY bias correction significantly improves cell clustering accuracy.

Conclusions:

• PATTY provides an effective solution for mitigating open-chromatin bias in CUT&Tag data.
• The method enhances the reliability of epigenomic profiling, particularly for sensitive single-cell analyses.
• PATTY offers a generalizable framework for bias correction in other Tn5 transposase-based genomic assays.