mRNALoc: a novel machine-learning based in-silico tool to predict mRNA subcellular localization.

Area of Science:

• Molecular Biology
• Bioinformatics
• Computational Biology

Background:

• Messenger RNA (mRNA) localization near translated proteins optimizes post-transcriptional gene expression.
• Nuclear mRNA retention regulates protein levels and temporal translation.
• mRNA localization is crucial for cellular functions including signaling and development.

Purpose of the Study:

• To develop a novel machine-learning tool, mRNALoc, for predicting subcellular locations of eukaryotic mRNAs.
• To utilize cDNA/mRNA sequences for accurate subcellular localization predictions.

Main Methods:

• Development of a machine-learning based tool named mRNALoc.
• Prediction of five subcellular locations: extracellular region, endoplasmic reticulum, cytoplasm, mitochondria, and nucleus.
• Validation using five-fold cross-validation and independent datasets.

Main Results:

• mRNALoc achieved high prediction accuracies across various subcellular locations.
• Cross-validation accuracies ranged from 65.19% to 99.70%.
• Independent dataset accuracies ranged from 58.10% to 96.88%, with AUC values up to 0.98.

Conclusions:

• mRNALoc is an effective tool for predicting eukaryotic mRNA subcellular localization.
• The tool facilitates research into mRNA localization's role in protein expression and cellular processes.
• mRNALoc is available as standalone software and a web server for academic use.