OGSA

OGSA assesses pathway deregulation in cancer subtypes by integrating outlier analysis across molecular data types and applying gene set analysis plus a top-scoring pair algorithm to identify deregulated pathways and associated biomarkers, including applications to pediatric acute myeloid leukemia (AML).

Key Features:

• Integration of Molecular Data Types: Combines outlier analysis within and across molecular data types such as gene mutations and epigenetic modifications to identify pathway-level signals.
• Outlier Identification across Molecular Alterations: Detects outliers in gene mutations, epigenetic modifications, overexpression, and gene amplifications/deletions.
• Gene Set Analysis: Evaluates the collective significance of pathway members to detect globally deregulated pathways rather than single-gene effects.
• Top-Scoring Pair Algorithm: Employs a top-scoring pair algorithm to identify robust biomarkers associated with pathway deregulation.
• Application in Pediatric AML: Applied and validated on pediatric acute myeloid leukemia datasets, including independent validation datasets and relapsed pediatric AML cases.

Scientific Applications:

• Pathway Deregulation Analysis in Cancer: Pinpoints pathways that are deregulated across cancer subtypes to inform mechanistic studies of carcinogenesis.
• Biomarker Discovery for Therapeutic Development: Identifies biomarkers linked to deregulated pathways to support drug development and therapeutic strategy design.
• Target Identification for Intervention: Aids in prioritizing pathway members and alterations as potential targets for intervention.
• Pediatric AML Subtype Characterization: Supports biomarker development and subgroup characterization for pediatric acute myeloid leukemia, including relapsed cases.

Methodology:

OGSA identifies outliers within molecular data types (gene mutations, epigenetic modifications, overexpression, gene amplifications/deletions), integrates these outliers with gene set analysis to detect significantly deregulated pathways, and applies a top-scoring pair algorithm to identify biomarkers associated with those pathways.