DeconPeaker

DeconPeaker deconvolutes cell type proportions in complex biological samples by integrating chromatin accessibility (ATAC-Seq) with gene expression data (RNA-Seq and microarray) to resolve cellular heterogeneity.

Key Features:

• Cell type estimation: Defines unique subpopulations within mixed samples using open chromatin (ATAC-Seq) patterns together with gene expression data.
• Performance metrics: Demonstrates low average root-mean-square error (RMSE = 0.042) and high average correlation coefficient (r = 0.919) compared with other deconvolution methods.
• Disease application (AML): Applied to acute myeloid leukemia chromatin accessibility data to identify cell types associated with disease progression.
• Chromatin accessibility indicator: Emphasizes that chromatin accessibility provides stronger distinguishing characteristics for cell-type identification than gene expression alone.
• Multi-omic integration: Combines chromatin accessibility and gene expression datasets to improve resolution of cellular composition.

Scientific Applications:

• Cancer research: Identification of specific cell types and tumor heterogeneity using ATAC-Seq and gene expression integration.
• Stem cell biology: Discrimination of subpopulations relevant to stem cell differentiation and niche dynamics.
• Immunology: Dissection of immune cell composition in mixed samples to inform studies of immune responses and disease.

Methodology:

Implements an algorithmic framework that integrates ATAC-Seq chromatin accessibility with RNA-Seq and microarray gene expression profiles to estimate cell-type proportions, reporting RMSE = 0.042 and average correlation r = 0.919.