This release adds Average Nucleotide Identity (ANI) as a dedicated analysis tool, rebuilds the dotplot viewer on Plotly, upgrades Gene Trait Matching with Scoary 2 phylogenetic correction, and extends background job processing to cover more tools across the application.
New Features
ANI - Average Nucleotide Identity
Arx now includes a dedicated genome similarity page with heatmap and dendrogram views. Two backends are supported: gendiscal (fast, recommended for most use cases) and pyani-plus (supports fastANI, ANIb, ANIm, and dnadiff). ANI jobs run fully in the background - start a calculation and come back when it's done. Results can be opened directly in the Trees tool to build a phylogenetic tree from the same genome set. Genome files (FASTA/GenBank) can also be uploaded directly on the ANI page for quick pairwise comparisons without adding genomes to the library first.
Dotplot Viewer
The dotplot viewer has been rebuilt on Plotly. Gene annotations from GenBank files are now rendered as arrows directly on the alignment, scaffold tracks are color-coded, and keyboard shortcuts (arrow keys, +/-) and double-click-to-zoom are supported. DNA bases are rendered directly on alignment tracks when zoomed in sufficiently. The new implementation is faster and supports bookmarkable URLs and a configurable settings panel.
Gene Trait Matching - Scoary 2
Gene Trait Matching now supports Scoary 2 with phylogenetic correction, which accounts for population structure and reduces false positives in closely related datasets. Jobs run asynchronously in the background, and results open in a new interactive Scoary Viewer for per-trait exploration. A new mGWAS (microbial GWAS) exploration page has also been added.
Job Queue Improvements
Job status panels have been fully redesigned across all tools. Jobs now show elapsed time, support pinning to keep important jobs visible, and receive automatic descriptive titles. BLAST and ANI jobs now appear on the central Job Queue page. Failed tree and ANI jobs are automatically retried.
Tree Building
GTR+G is now the default substitution model for IQ-TREE, improving phylogenetic accuracy. WAG+G has been added as an option. ClustalO is now bundled - no separate installation required. Trees can also be uploaded directly on the trees page.
Home Page & Navigation
The home page sunburst chart now supports filtering and searching by genome attributes. Selected genomes and BLAST inputs are remembered across page reloads using session storage. Custom navigation links can be added to the navbar via the NAV_LINKS environment variable.
Genome Viewer
- Contig table is now scrollable.
- Improved click menus with better intron handling.
- Improved file browser.
Genome Upload and Import
- Prokka placeholder sequences are detected on upload and flagged before import.
- Improved chromosome detection and strain fallback logic.
- Friendlier taxid error messages and clearer gene count mismatch errors.
- Advanced option to skip sanity checks during folder structure import.
Infrastructure & Backend
- Folder Structure V3 is now supported. Genome files are normalized to use consistent, genome-identifier-based contig IDs and locus tags.
- Updated OrthoFinder to the latest version.
- Improved container startup reliability and nginx fallback handling.
- New postgres-benchmark management command to measure database connection latency, write throughput, and query speed.
- PostgreSQL JIT disabled globally for improved query stability.
Performance
- Reduced redundant database queries in the DNA viewer and tools.
- Pathway tool loads significantly faster.
- Improved annotation import pipeline and search performance.
Bug Fixes
- Fix gene and annotation rendering in DNA viewer.
- Fix click menus for dotplot and trees.
- Fix column filter in tables.
- Fix backspace behavior in magic tag autocomplete.
- Fix BLAST taxonomy display.
- Fix dotplot URL not restoring correctly on reload.
- Fix matplotlib SVG backend resource leak.