iMIRAGE

iMIRAGE imputes microRNA (miRNA) activity from protein-coding gene (PCG) expression using machine-learning algorithms to enable miRNA profiling from transcriptomic datasets.

Key Features:

• R package: Implements the methods as an R package for integration into R-based transcriptomic analyses.
• Machine-learning imputation: Uses machine-learning algorithms to predict miRNA expression indirectly from PCG expression data.
• Normalization and transformation: Provides an integrated workflow for normalization and transformation of both miRNA and PCG expression data.
• Predicted miRNA targets: Offers an option to use predicted miRNA targets as the basis for imputing miRNA activity from independent PCG datasets.
• Regulatory inference: Leverages miRNA-mediated effects on PCG expression to infer miRNA activity.

Scientific Applications:

• Imputation in datasets lacking small RNA profiling: Imputes miRNA activity in transcriptomic datasets that lack reliable small RNA profiling.
• Comparative miRNA activity analysis: Enables comparison of inferred miRNA activity across phenotypes to study miRNA-mediated regulatory mechanisms.
• Retrospective public dataset analysis: Allows incorporation of inferred miRNA profiles into analyses of publicly available transcriptomes that lack measured miRNA data.

Methodology:

Normalization and transformation of miRNA and PCG expression followed by machine-learning-based imputation of miRNA activity from PCG expression, optionally using predicted miRNA targets.