SigCheck

SigCheck evaluates gene signatures by comparing them to random gene sets, known related and unrelated signatures, permuted data and metadata to assess their performance in survival analysis and classification tasks.

Key Features:

• Performance Evaluation: Compares an input gene signature to random gene sets of equivalent length, known related and unrelated signatures, and permuted datasets and metadata to quantify robustness and reliability.
• Survival and Classification Analysis: Evaluates signatures in the contexts of survival analysis and classification tasks to measure predictive performance.
• Input Format: Accepts ExpressionSet-formatted high-throughput genomics data as input.
• Implementation: Performs statistical analyses within the R programming environment and integrates with Bioconductor.

Scientific Applications:

• Phenotype Prediction: Assesses gene signatures for their ability to predict phenotypes from genomic datasets.
• Validation of Gene Signatures: Provides systematic benchmarking to validate the relevance and effectiveness of gene signatures in genomic studies.

Methodology:

Accepts ExpressionSet input; compares the input signature against multiple benchmarks including random gene sets of equivalent length, known related and unrelated signatures, and permuted datasets and metadata; and applies statistical analyses in R and Bioconductor.