Volume II: Methods, Tools and Algorithms in Single Cell Bioinformatics

Over the last few years, the single-cell landscape has transitioned from mapping individual cellular transcriptomes to generating atlas-scale, spatially resolved, and multimodal datasets. While these advancements open unprecedented views into tissue architecture and cellular dynamics, they also introduce significant computational bottlenecks. Volume II of this Research Topic captures the next generation of methodological and algorithmic innovations required to process, model, and translate these complex data structures into biological insights.We invite submissions presenting innovative tools, robust mathematical methods, and scalable algorithms, with a particular focus on four key areas:• AI, Foundation Models, and Predictive Cell Modeling Novel machine learning architectures — including transformer-based and deep generative models — for cell-type annotation, transfer learning across datasets, zero-shot prediction, and low-dimensional representation of heterogeneous cellular states. Topics include cell foundation models, perturbation prediction, in silico drug response, digital twins, generative biology, synthetic perturbation modeling, and cross-species transfer learning.• Next-Generation Multimodal and Spatial Integration Advanced computational frameworks for aligning multi-omic layers (e.g., scRNA-seq, scATAC-seq, proteomic, and epigenomic data) and modeling high-resolution spatial transcriptomics, spatial proteomics, and temporal lineage tracing. Topics include ligand-receptor inference, probabilistic communication models, graph neural networks, network biology, immune ecosystems and niches, spatial and multicellular communication, and mechanistic signaling models.• Atlas-Scale Computations, Benchmarking, and Clinical Translation High-performance streaming algorithms, distributed computing frameworks, uncertainty quantification, and rigorous benchmarking protocols. We strongly encourage user-facing, interoperable software pipelines built under FAIR principles (e.g., Nextflow, Snakemake, nf-core, CWL, Docker, Singularity) that facilitate clinical diagnostics, therapeutic target discovery, and patient stratification — including biomarker discovery, treatment response prediction, precision medicine, digital pathology, companion diagnostics, and regulatory-grade pipelines• Statistical Inference, Causal Modeling, and Interpretable AI: While foundation models and deep learning excel at complex pattern recognition, rigorous statis-tical frameworks remain paramount for hypothesis testing, causal inference, and biological interpretability. We welcome submissions that develop advanced high-dimensional statistical techniques to tackle inherent single-cell noise, sparsity, and complex batch effects. We particularly encourage comparative studies evaluating traditional statistical methods against modern artificial intel-ligence, as well as hybrid approaches that integrate the interpretability of prob-abilistic models with the predictive power of machine learning. This second volume continues to champion transparent validation, open-source code availability, and community standards. Submissions must demonstrate a clear advancement over existing methodologies through comparative evaluation on benchmark datasets. Topic editors: Xinru Qiu, Maciej Pietrzak, Abhijeet R Patil, Edgar Gonzalez-Kozlova, Nikolaos Patikas Source: https://www.frontiersin.org/research-topics/83924/volume-ii-methods-tools-and-algorithms-in-single-cell-bioinformatics