DDR

DDR identifies platform-independent gene expression biomarkers and performs single-sample classification by using stably expressed housekeeping genes to mitigate platform-specific biases in high-throughput expression data.

Key Features:

• Platform-Independent Biomarker Identification: Utilizes stably expressed housekeeping genes as references to mitigate platform-specific biases and non-biological variability across profiling technologies.
• Robust Cross-Platform Gene Signature: Identifies small, reproducible gene signatures for per-patient disease classification by harmonizing data from diverse expression sources.
• Validation Across Technologies: Validated on RNA-seq data from blood platelets to classify tumor types and molecular target statuses (MET, HER2-positive, mutant KRAS, EGFR, PIK3CA) using smaller biomarker sets than traditional methods.
• Application to Microarray Data: Applied to three microarray datasets to identify robust biomarkers for medulloblastoma subgrouping, recovering known subgroup-specific markers and detecting potential new biomarkers.
• Contribution to Precision Medicine: Enables cross-platform diagnostic and prognostic gene signatures for single-patient disease classification to support individualized treatment strategies.

Scientific Applications:

• Biomarker discovery for diagnosis and prognosis: Supports identification of expression biomarkers relevant to early diagnosis and prognostic assessment.
• Single-sample disease classification: Provides per-patient classification independent of underlying profiling platforms.
• Tumor type and molecular target classification: Classifies tumor types and molecular target statuses (MET, HER2-positive, mutant KRAS, EGFR, PIK3CA) using blood platelet RNA-seq signatures.
• Medulloblastoma subgrouping: Identifies known and candidate subgroup-specific biomarkers across microarray platforms for medulloblastoma sample stratification.
• Cross-platform transcriptome signature development: Generates gene signatures that are reproducible across RNA-seq and microarray technologies for precision-medicine applications.

Methodology:

Uses stably expressed housekeeping genes as reference controls in high-throughput gene expression profiles to identify cross-platform gene signatures and validates those signatures on RNA-seq blood platelet data and three microarray datasets.