Semi-parametric Empirical Bayes Method for Multiplet Detection in snATAC-seq with Probabilistic Multi-omic...

Area of Science:

• Genomics
• Computational Biology
• Bioinformatics

Background:

• Multiplets, where multiple cells are captured in a single droplet, create hybrid molecular profiles that complicate single-cell analyses.
• Detecting multiplets in single-nucleus ATAC-seq (snATAC-seq) data is difficult due to sparse and overdispersed chromatin accessibility measurements.

Purpose of the Study:

• To introduce SEBULA, a novel semi-parametric empirical Bayes model for accurate multiplet detection in snATAC-seq data.
• To enable principled false discovery rate control in multiplet identification.
• To integrate multimodal data, including scRNA-seq, for enhanced multiplet detection.

Main Methods:

• Development of SEBULA, a semi-parametric empirical Bayes model.
• Utilizing well-calibrated posterior probabilities for multiplet detection.
• Integration of probabilistic evidence with complementary signals from other data modalities (e.g., scRNA-seq).

Main Results:

• SEBULA provides well-calibrated posterior probabilities for multiplet detection.
• The model enables principled control over the false discovery rate.
• Benchmarking on simulations and seven annotated trimodal DOGMA-seq datasets confirmed SEBULA's superior performance.
• The open-source SEBULA software is computationally efficient.

Conclusions:

• SEBULA effectively addresses the challenge of multiplet detection in snATAC-seq data.
• The model enhances the reliability of single-cell analyses by accurately identifying and controlling for multiplets.
• SEBULA offers a computationally efficient and robust solution for multiplet detection, with potential applications in multi-modal single-cell studies.