massiR

massiR predicts sample sex in gene expression microarray datasets to enable accurate sex-specific transcriptome analyses and meta-analyses.

Key Features:

• Sex prediction using Y-chromosome probes: Uses probe-level expression data from genes on the Y chromosome to distinguish male from female samples.
• Unsupervised clustering: Applies unsupervised clustering techniques to group samples by presence or absence of Y-linked signal for sex classification.
• Implementation: Provided as an R package within the Bioconductor project for integration with R-based bioinformatics workflows.
• Label quality control: Identifies potential sample mislabeling or annotation errors through discordance between predicted and recorded sex.

Scientific Applications:

• Resolving missing sex metadata: Infers sex where sample annotation is absent or incomplete in microarray datasets.
• Sex-stratified transcriptome analysis: Enables incorporation of sex as a biological variable in differential expression and downstream analyses.
• Meta-analysis harmonization: Improves comparability of datasets in meta-analyses by providing consistent sex labels.
• Mammalian transcriptomics quality control: Supports detection of sex-specific expression patterns and dataset integrity checks in mammalian studies.

Methodology:

The method analyzes microarray probe data targeting Y chromosome genes and applies unsupervised clustering to classify samples as male or female based on the presence or absence of Y-linked probe expression, enabling detection of mislabeled samples.