ADC

ADC computes approximate distance correlation to identify genes that are statistically interrelated across distinct biological omics datasets by quantifying linear and nonlinear associations for the same gene between two datasets.

Key Features:

• Cross-Dataset Gene Relationship Analysis: Quantifies the strength of association for the same gene across two distinct biological datasets derived from different samples, enabling detection of interrelations beyond single-sample correlations.
• Approximate Observations: For each target gene, identifies the k most correlated genes within each dataset to form approximate observations used for downstream association estimation.
• Distance Correlation Calculation: Computes distance correlation (DC) for each target gene across the two datasets to capture both linear and nonlinear associations.
• Statistical Significance and FDR Control: Applies the Benjamini-Hochberg procedure to adjust p-values for multiple testing and control the false discovery rate (FDR).
• Scalability and Efficiency: Implements scalable computational strategies suitable for large-scale omics datasets.

Scientific Applications:

• Cancer Research: Analysis of 21 RNA-seq datasets from different tissues to reveal interrelated genes with potential implications in cancer biology.
• Hematopoietic Cell Lineage Studies: Applied to six single-cell RNA-seq (scRNA-seq) datasets of mouse hematopoietic cells to identify gene associations across six cell types.
• Pancreatic Islet Cell Research: Across five scRNA-seq datasets from pancreatic islet cells generated using different technologies to identify consistent interrelated genes.
• Integrated Multi-Omics Analysis: Analyzed coupled single-cell ATAC-seq (scATAC-seq) and scRNA-seq data of peripheral blood mononuclear cells (PBMC) to explore gene regulation relationships across modalities.

Methodology:

For each target gene, ADC forms approximate observations by selecting the k most correlated genes within each dataset, computes distance correlation (DC) between the two datasets for that gene, and applies the Benjamini-Hochberg procedure to adjust for multiple testing and control FDR.