ADAPTS

ADAPTS performs tissue-specific cell-type deconvolution of gene expression data to infer immune cell presence and proportions in heterogeneous tumor samples for tumor immunology studies.

Key Features:

• Signature Matrix Customization: Customizes or constructs signature matrices from purified cell-type gene expression profiles, including single-cell RNA sequencing (scRNA-seq) data.
• Integration with Deconvolution Algorithms: Integrates signature matrices with multiple deconvolution algorithms to estimate proportions of different cell types in mixed samples.
• Hierarchical Deconvolution Technique: Clusters cell types that are difficult to distinguish and applies hierarchical deconvolution to re-split these clusters for improved resolution.
• Improved Predictive Analysis: Uses blind predictive analysis to improve reconstruction of cell types in single-cell RNA-seq datasets.

Scientific Applications:

• Tumor Immune Infiltration Analysis: Infers immune cell infiltration patterns in tumors to support tumor immunology research.
• Patient Outcome Association: Associates inferred immune cell proportions with patient prognosis and treatment response.
• Single-cell RNA-seq Reconstruction: Enhances reconstruction and interpretation of cell types within single-cell RNA-seq datasets.

Methodology:

Derives signature matrices from purified cell-type gene expression data (including scRNA-seq), uses those matrices as references for deconvolution algorithms to estimate cell-type proportions, applies clustering and hierarchical deconvolution to resolve ambiguous cell-type groups, and employs blind predictive analysis for reconstruction.