puma

"puma" (Propagating Uncertainties in Microarray Analysis) has been significantly extended and improved to address the evolving needs of gene and transcript expression analysis, particularly with the rising interest in isoform-level expression profiling. Initially designed for Affymetrix 3' GeneChip data, puma now includes advanced methods tailored for GeneChip Exon arrays, offering an essential tool for quantifying expression levels of most exons and measuring isoform-level expression.

Key enhancements introduced in the updated version of puma include: 1. A novel gamma model specifically designed for exon arrays calculates isoform and gene expression measurements and assesses the level of uncertainty associated with these estimates. This is achieved by leveraging the multi-mappings between probes, isoforms, and genes. 2. A refined approach for probe-level analysis of Affymetrix 3' GeneChip data, aimed at generating more stable gene expression estimates, particularly for low expression genes. 3. An improved method for detecting differential expression, which is computationally more efficient compared to the previous approach within puma, facilitating faster analysis without compromising accuracy. 4. An enhanced model for robust, model-based clustering of gene expression that incorporates variance from technical and biological replicates, enabling more reliable identification of expression patterns.

These extensions and improvements make puma a comprehensive and versatile package for analyzing Affymetrix 3' GeneChips and Exon arrays. By effectively propagating the uncertainty of expression measurements, puma enables more efficient and thorough downstream analysis at both gene and isoform levels. The downstream methods also apply to other expression quantification platforms, such as RNA-Seq, whenever uncertainty information is available from expression measurements.