Advancing Interdisciplinary Research using Galaxy

Galaxy is an open-source, web-based platform designed to empower researchers by providing an accessible and powerful environment for data analysis, management, and sharing. Initially developed for genomics, Galaxy has expanded into a versatile computational workbench supporting diverse domains, including single-cell genomics, biodiversity, microbiome research, clinical data analysis, ecology, climate science, and humanities. Its intuitive interface allows users, regardless of computational expertise, to upload datasets, execute complex analytical workflows, and visualize results, all while maintaining detailed provenance tracking for reproducibility. Galaxy has been contributing to the emerging domains of single-cell and spatial omics with integration of a wide array of tools and workflows tailored for single-cell RNA-seq, ATAC-seq, multiomics and spatial transcriptomics. Additionally, Galaxy features advanced workflows built on widely adopted tool sets such as Scanpy, Seurat, and the broader scverse ecosystem. These analytical capabilities are further supported by rich, community-curated training resources and hands-on tutorials developed by the Galaxy Training Network (GTN), which provide step-by-step guidance using real-world datasets. In biodiversity, Galaxy supports large-scale genomics via the Vertebrate Genomes Project (VGP) and the European Reference Genome Atlas (ERGA) within the Biodiversity Genomics Europe (BGE) framework. VGP uses Galaxy's standardized workflows to generate reference-quality genomes for 74,963 vertebrate species, ensuring consistency, transparency, and open access through repositories like Genome Ark. ERGA employs Galaxy to sequence ~200,000 European eukaryotic species, aligning with Earth BioGenome Project (EBP) standards for high-quality genomes to enhance biodiversity monitoring and conservation. ERGA-BGE researchers have developed specialized workflows, such as the ONT+Illumina & HiC pipeline, integrating long/short reads and Hi-C data for robust assemblies, shared via WorkflowHub, empowering global genomics research. Considerable progress has been made on several fronts that together turn Galaxy into a platform suitable for the analysis of sensitive human data. Galaxy has enhanced its capability for sensitive human data analysis by ensuring user-uploaded data remains inaccessible to others. Users can configure remote storage under their control for data and analysis results, which becomes fully inaccessible when disconnected from Galaxy. Users can also connect their own compute nodes via Pulsar, a distributed job execution server for Galaxy. The European Galaxy server is now ISO-27001-certified, and ongoing collaborations with European Genome Archive nodes support crypt4gh-standard encrypted data management, decrypted only on compute nodes for immediate analysis. Galaxy also offers flexible on-premises deployment, including Docker and Kubernetes, enabling hospitals to use in-house Galaxy deployments for clinical research and patient care data analysis. In microbiome research, the microGalaxy Special Interest Group (SIG), aims to develop and sustain microbial data analysis in Galaxy by establishing best practices, expanding training resources, coordinating collaborative efforts to prioritize tools and workflows, and engaging microbiology research communities. It maintains over 280 tool suites for tasks like taxonomic profiling, assembly, annotation, and antimicrobial resistance detection, alongside 50+ customizable workflows and 30+ tutorials accessible via the Galaxy Training Network (GTN), fostering standardized, community-driven microbiome research. Beyond genomics, Galaxy supports humanities research by integrating large language models (LLMs) and AI tools. For instance, NFDI4Memory demonstrated that ChatGPT's effectiveness in Named Entity Recognition (NER). Galaxy now includes a GPT interface (requiring a personal API key) and Whisper's for multilingual audio-to-text conversion. Researchers can upload .MP3 files, transcribe them, apply GPT for NER, and map results in QGIS, adapting workflows to diverse academic needs.