In hashtag#RNA research, one of the biggest challenges lies in how we transform raw RNA sequences into meaningful features that hashtag#machine_learning models can understand. This process, known as hashtag#feature_extraction, is the foundation of predictive modelling for RNA classification, interaction analysis, and functional characterization.
🔎 Yet, despite its importance, the field has remained broad, fragmented, and lacking standardization, making it difficult for researchers to compare studies, reproduce results, and build upon one another’s work.
✨ To address this gap, our review article brings clarity and structure by systematically consolidating more than 25 feature extraction strategies into sequence-based and structure-based approaches, providing a comparative analysis of how different feature sets influence model performance, and highlighting practical applications with a curated list of publicly available tools and packages to lower the barrier for adoption.
Our aim is to empower researchers with a clear roadmap for integrating feature extraction into RNA machine learning studies, enhancing reproducibility, scalability, and interpretability.
📖 You can read the full preprint here
Fatemeh Vafaee
Vafaee Lab