MiRanalyzer

MiRanalyzer detects and characterizes microRNAs from next-generation sequencing (NGS) small RNA datasets, identifying known miRNAs and predicting novel miRNA candidates for biological interpretation.

Key Features:

• Input Requirements: Accepts input of unique reads with corresponding copy numbers representing expression levels.
• Detection and Identification: Detects known microRNA sequences annotated in miRBase and identifies perfect matches against other libraries of transcribed sequences.
• Prediction of Novel miRNAs: Predicts novel microRNAs using a machine learning algorithm that reported an AUC of 97.9% and recall up to 75% on unseen data.

Scientific Applications:

• Known miRNA identification: Identification of known miRNA sequences within large NGS-derived small RNA datasets.
• Novel miRNA discovery: Discovery of novel miRNAs across species using predictive models, including species with limited annotated miRNAs.
• Small RNA sequencing analysis: Analysis and interpretation of small RNA sequencing data to assess miRNA expression and composition.

Methodology:

Performs sequence alignment and matching to annotated databases (miRBase) and other transcribed sequence libraries, and applies advanced machine learning algorithms for novel miRNA prediction.