SPD

SPD infers sample progression and identifies gene subsets from microarray gene expression data to reveal underlying biological processes such as differentiation, development, cell cycle, and disease progression.

Key Features:

• Unsupervised discovery: Operates without prior information about sample temporal order or gene functions, performing unsupervised inference from microarray gene expression data.
• Biological progression mapping: Organizes samples along an inferred progression continuum to reveal temporal dynamics of biological systems.
• Gene identification: Identifies subsets of genes associated with the inferred progression and that may drive the observed biological process.
• Versatility across biological processes: Applied to cell cycle time series, B-cell differentiation, mouse embryonic stem cell (ESC) differentiation, and prostate cancer progression, recovering sample order and relevant genes in each case.
• Hypothesis synthesis: Provides an inferred progression model and candidate regulatory genes to support generation of biological hypotheses.

Scientific Applications:

• Cell cycle time series: Recovers sample ordering and implicates progression-associated genes in cell cycle microarray datasets.
• B-cell differentiation: Infers differentiation order and identifies genes associated with B-cell development.
• Mouse embryonic stem cell (ESC) differentiation: Recovers differentiation stages and relevant genes in mouse ESC datasets.
• Prostate cancer progression: Infers disease progression order and identifies associated genes in prostate cancer microarray datasets.
• Hypothesis generation in genomics and systems biology: Synthesizes progression models and candidate regulatory genes to generate testable hypotheses.

Methodology:

Unsupervised inference of sample ordering and selection of progression-associated gene subsets directly from microarray gene expression data without requiring prior sample temporal order or predefined gene lists.