ADTnorm Publication Accepted by Nature Communications

We are thrilled to announce that our paper on ADTnorm has been accepted for publication in Nature Communications. This work represents a significant advancement in the normalization and integration of CITE-seq data.

A Breakthrough in CITE-seq Data Analysis

The publication, titled “ADTnorm: Normalization and integration of antibody-derived tag abundance in CITE-seq data,” introduces a novel computational method specifically designed for processing antibody-derived tags (ADTs) in CITE-seq experiments.

CITE-seq (Cellular Indexing of Transcriptomes and Epitopes by Sequencing) enables the simultaneous measurement of surface protein and mRNA expression in single cells. While this technology offers unprecedented insights into cellular biology, technical variability in antibody staining creates significant batch effects that complicate data interpretation and cross-study analyses.

Key Features of ADTnorm

ADTnorm addresses these challenges with several innovative approaches:

• Effective removal of technical variation across batches
• Accurate alignment of negative and positive protein marker expressions
• Improved cell-type separation
• Built-in utilities for automated threshold-gating
• Tools for assessing antibody staining quality for titration optimization
• Support for antibody panel selection

Benchmarking Excellence

The team rigorously benchmarked ADTnorm against 14 existing scaling and normalization methods across 13 public datasets. The results demonstrate ADTnorm’s superior performance in aligning populations with negative and positive expression of surface protein markers.

Real-World Applications

Beyond its technical achievements, ADTnorm has demonstrated significant practical value. When applied to a published COVID-19 CITE-seq dataset, the tool identified previously undetected disease-associated markers, highlighting its broad utility in biological applications.

Availability and Access

The software is freely available on GitHub: ADTnorm Repository

The paper preprint can be accessed at: bioRxiv

Processed data from the 13 public studies used in the analysis are available as part of the ADTnorm software repository: Demo Data

Acknowledgments

We would like to thank all collaborators and contributors who made this work possible. This publication represents a significant milestone in our ongoing efforts to develop innovative computational methods for biomedical research.