panp

panp performs presence–absence calls using negative strand matching probesets (NSMP) to determine gene presence or absence from Affymetrix microarray data.

Key Features:

• Negative strand matching probesets (NSMP): Uses NSMP derived from Affymetrix annotations to inform presence/absence calls.
• Decoupled processing: Operates independently of the preprocessing method that generates gene expression values, supporting PM-only methods like RMA and PM/MM methods such as MAS5 and GCRMA.
• Supported chipsets: Provides NSMP sets for HGU133A and HGU133-Plus-2.0 Affymetrix GeneChip arrays.
• Improved detection accuracy: Demonstrates superior probeset detection compared to the MAS5 presence-absence (MAS-P/A) method, showing higher accuracy and precision across a full range of transcript concentrations including low-abundance transcripts.

Scientific Applications:

• Gene presence/absence calling: Determines presence or absence of genes from Affymetrix microarray experiments using NSMP on HGU133A and HGU133-Plus-2.0 arrays.
• Spike-in benchmarking: Benchmarks detection performance using the Affymetrix Latin squares spike-in dataset consisting of 28 HG-U133A chips.
• TaqMan validation: Validates presence/absence calls against a TaqMan-validated human tissue dataset on HG-U133 Plus 2.0 chipsets.
• Method comparison: Evaluates and compares performance relative to MAS5 presence-absence calls across datasets and preprocessing methods.

Methodology:

Analyzes NSMP data using statistical techniques implemented in R and decouples gene expression value generation from presence/absence detection calls.