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

Abstract

Accurate epigenomic profiling is essential for understanding chromatin and gene regulation. Cleavage Under Targets & Tagmentation (CUT&Tag) enables efficient epigenomic profiling for low-input and single-cell samples and is increasingly adopted in genomics research. However, CUT&Tag's usage of Tn5 transposase introduces bias toward open chromatin that distorts genome-wide signal patterns and confounds downstream analysis, especially in sparse single-cell data. We show that open-chromatin bias extensively exists in published CUT&Tag data, including those generated with optimized high-salt protocols. To address this issue, we present PATTY (Propensity Analyzer for Tn5 Transposase Yielded bias), a computational method that corrects open-chromatin bias in CUT&Tag by leveraging accompanying ATAC-seq using machine learning. PATTY enables accurate and robust signal detection for both active and repressive histone modifications, including H3K27ac, H3K27me3, and H3K9me3. In single-cell CUT&Tag analyses, PATTY bias correction improves cell clustering accuracy. Beyond CUT&Tag, PATTY establishes a generalizable framework for bias correction in Tn5-based genomic assays.