Decoding RNA Biology with Artificial Intelligence: From Single-Cell Transcriptomics to Non-Coding RNA Networks

The transcriptomic landscape of living cells is defined not only by the genes that are expressed, but by the intricate networks of coding and non-coding RNAs that govern when, where, and how that expression is regulated. Disentangling this complexity demands analytical tools that can match its scale, and artificial intelligence is increasingly providing them.Single-cell transcriptomics has transformed our ability to resolve cellular identity, state transitions, and population dynamics at unprecedented resolution. Yet the analytical challenges are substantial, dropout events, batch effects, high dimensionality, and the need to integrate multi-modal data across time and space. AI and machine learning approaches, from graph neural networks and transformer architectures to generative models, are emerging as the most powerful means of addressing these challenges, enabling robust cell type classification, trajectory inference, spatial transcriptomic integration, and cross-dataset harmonization.Non-coding RNA networks, spanning long non-coding RNAs, microRNAs, circular RNAs, and small nucleolar RNAs, form a regulatory layer of enormous functional depth. Their combinatorial interactions with RNA-binding proteins, chromatin, and the translational machinery create networks whose logic is only beginning to be decoded. AI-driven approaches, including network inference algorithms, large language models trained on RNA sequences, and deep learning-based structure prediction, are opening new windows into ncRNA function, target prediction, and their roles in disease.The synthesis of these two pillars, AI applied to single-cell data and to non-coding RNA networks, defines a cohesive and highly productive research space that aligns directly with the scope of this section. Contributions are invited that demonstrate the power of AI to reveal RNA biology’s regulatory architecture, whether through new computational frameworks, benchmark studies comparing methodological approaches, or biological discoveries enabled by these methods.This Research Topic welcomes submissions on:- Machine learning and deep learning for single-cell RNA-seq analysis (clustering, trajectory, imputation, integration)- Spatial transcriptomics and multi-modal data fusion- AI-based prediction of non-coding RNA structure, targets, and regulatory function- Network inference and graph-based models of RNA regulatory systems- Large language models and foundation models applied to RNA sequences or transcriptomic data- AI tools for splicing regulation, RNA-protein interaction, and RNA stability- Clinical and disease applications of AI-driven transcriptomic and ncRNA analysesGiven the propensity of AI methods to identify or generate apparent patterns that may lack biological relevance, submissions are expected to support AI-based predictions with robust underlying data and, wherever feasible, independent validation, whether through internal biological markers or an orthogonal experimental/computational approach. Topic editors: Luca Alessandri, Roberto Pagliarini, Amit Kumar Pandey Source: https://www.frontiersin.org/research-topics/82483/decoding-rna-biology-with-artificial-intelligence-from-single-cell-transcriptomics-to-non-coding-rna-networks