VX: an AI-agent-controllable desktop genome and transcriptome browser with a programmable analysis framework

A fast, cross-platform, GPU-accelerated desktop genome and transcriptome browser, built from the ground up for AI-agent driven exploration of sequencing data. Visualise FASTA / GTF / GFF / BAM / BigWig / BigBed / VCF / Hi-C interaction tracks; run 50+ built-in computational analyses against them; drive the whole thing from Claude, an LLM agent, or a plain curl command through an embedded HTTP API.

---

What VX is for

VX is a desktop genome browser in the family of IGV, UCSC Genome Browser, JBrowse, Jbrowse 2, SeqMonk, Tablet, Pileup.js, and Pretzel, but built for a different decade. It targets four audiences that the current generation of browsers serves less well:

1. Researchers working with very large sequencing datasets (multi-GB BAM, BigWig, Hi-C .cool/.mcool, multi-sample VCF) on a laptop, who need an instantly responsive viewer that doesn't lock up while a 23 GB BAM region cache warms.
2. Bioinformaticians and computational biologists who want to run analyses against the data while looking at it. Run peak calling, GC content, coverage anomalies, alignment QC, variant density, Hi-C TAD boundaries, cross-track correlation, directly from the viewer, in the background, with file exports.
3. AI / agentic-LLM users who want a genome browser their agent can see and control: navigate the viewport, load files, screenshot tracks, run analyses, export regions, all through a structured tool catalogue.
4. Educators and figure-makers who want a reproducible, scriptable, screenshot-and-record-ready viewer that produces publication-grade output (PNG, SVG, MP4 / WebM / GIF screen recordings) from a single binary.

If you do NGS, RNA-seq, ChIP-seq, ATAC-seq, Hi-C, long-read sequencing, variant calling, comparative genomics, transcript isoform analysis, ribosome profiling, CLIP-seq, methylation sequencing, MeRIP-seq, or any other sequencing-based assay and you've ever wished your genome browser was faster, smarter, or scriptable, VX is built for you.

---

Why VX is different

Capability	What it means in practice
GPU-accelerated rendering	OpenGL-based level-of-detail engine: render tens of thousands of features per second, smooth pan/zoom from chromosome scale (10⁸ bp) to base level (10⁰ bp) in one continuous gesture. No tile-popping. No re-fetch on every nudge.
Async, background-thread BAM loading	The viewport never blocks. Load a 23 GB BAM, click to a new region, render proceeds at full FPS while the read fetch completes in the background. A load-status strip tells you why the alignments aren't there yet.
43-tool MCP catalogue + HTTP API	Built-in HTTP server on 127.0.0.1:9876. Any tool that speaks HTTP (Claude Code, Claude Desktop, a Python script, curl) can drive every aspect of the viewer: navigate, load files, query genes, screenshot the viewport, run analyses, manage bookmarks, record video. AI-native by design.
50+ built-in analyses	Peak calling, RPKM/TPM/CPM quantification, GC content, CpG density, motif search, repeat density, signal correlation, variant annotation, Ts/Tv, alleles, MAPQ distribution, insert sizes, strand bias, soft-clip frequency, Hi-C contact-vs-distance, TAD boundaries, heatmap matrix, profile plots, peak prominence, gap detection, all in a background thread with file output (BED / BedGraph / TSV).
Built-in video / GIF recording	Capture the OpenGL viewport to MP4 (H.264), WebM (VP9) or GIF. Two modes: timer-paced (constant fps) or render-paced (every frame). Quality and frame rate user-controlled. Ideal for figure animations, talks, supplementary movies.
Subsampling + transparency	Alignment tracks honour a configurable read-cap (default 500 reads/region) with a min_mapq filter. When a region is subsampled, a badge on the track tells you so. The screenshot endpoint respects the same cap, so all figures match what you see.
Self-contained binaries	No Python runtime, no Conda environment, no JBrowse-style web stack. A single binary (~19 MB on Linux, ~45 MB macOS, ~46 MB Windows zip) with all GUI deps either bundled (Win/Mac) or relying on a baseline desktop install (Linux). Drop, double-click, run.
Triple-platform pre-compiled	Linux x86_64 (4 CPU-profile binaries), macOS arm64 .app + .dmg, Windows x86_64 ZIP. Built with the LDC LLVM-based D compiler, full LTO, static druntime + Phobos.
No telemetry, no cloud, no account	VX is a desktop application. It does not phone home. The HTTP API binds only to 127.0.0.1 by default. Your data stays on your machine.

If you've used IGV and wished it were faster, or JBrowse and wished it weren't a browser tab, or Pileup.js and wished it ran big files, or any genome browser and wished an AI agent could drive it, VX is the answer.

---

Screenshots

Chromosome-scale overview — 1 ideogram + multiple tracks rendered at 10⁸ bp resolution.	Gene boundaries with intron-collapse for fast structure inspection.
Gene-level detail with transcript isoforms.	Exon-level detail showing UTRs, CDS, splice sites.
Base-level resolution with codon highlighting and frame translation.	Dark mode for long sessions / projector-friendly figures.
Multi-track composition: sequence + genes + coverage + alignments + signal + variants.	Anatomy of a gene track — UTRs, exons, introns, strand arrows.

---

Quick install

1. Download the right binary for your platform

Visit the Releases page and grab one asset for your system:

Platform	Asset	Size
Linux x86_64 (Haswell-class or newer; AVX2/BMI2/FMA)	VX-X.Y.Z-linux-x86_64.tar.gz	~6.6 MB
macOS arm64 (Apple Silicon, macOS 13+)	VX-X.Y.Z-macos-arm64.dmg	~18 MB
Windows x86_64 (Windows 10 22H2 or Windows 11)	vx-X.Y.Z-windows-x86_64.zip	~46 MB

Each asset has a matching .sha256 sidecar. Verify before running:

sha256sum    -c VX-X.Y.Z-linux-x86_64.tar.gz.sha256        # Linux
shasum -a 256 -c VX-X.Y.Z-macos-arm64.dmg.sha256           # macOS
certutil -hashfile vx-X.Y.Z-windows-x86_64.zip SHA256      # Windows

2. Unpack / install

Linux relocatable tarball:

tar xzf VX-X.Y.Z-linux-x86_64.tar.gz
cd vx-linux/
./bin/genome-viewer

Linux dependencies are expected to be present from a baseline GNOME / KDE / XFCE desktop install: glibc 2.28+, GTK 3.24+, GLib 2.60+, Cairo, Pango, freetype, libepoxy, an OpenGL driver (libGL.so.1).

• Ubuntu 20.04+: sudo apt install libgtk-3-0 libgl1 libepoxy0
• Fedora 35+ / RHEL 9: sudo dnf install gtk3 mesa-libGL libepoxy
• Arch / Manjaro: sudo pacman -S gtk3 libepoxy mesa

macOS arm64 .dmg installer:

1. Double-click VX-X.Y.Z-macos-arm64.dmg.
2. Drag VX.app into the Applications folder.
3. First launch: right-click VX.app → Open (Gatekeeper one-time confirmation; the bundle is ad-hoc signed only, no Developer ID).
4. Or, from Terminal: xattr -dr com.apple.quarantine /Applications/VX.app

Windows x86_64 relocatable ZIP:

1. Extract vx-X.Y.Z-windows-x86_64.zip anywhere.
2. Double-click genome-viewer.exe inside the extracted folder.
3. Windows SmartScreen may warn on first launch (unsigned). Click More info → Run anyway.

3. Verify it works

VX exposes an HTTP API on 127.0.0.1:9876. From any shell:

curl -s http://127.0.0.1:9876/ping
# → {"status":"ok","app":"VX Genome Viewer","api_version":2}

That's your installation test.

---

Drive VX from Claude or any LLM (MCP)

VX ships with an open-source MCP server (mcp/vx_mcp_server.py, MIT) that exposes the entire viewer to AI agents that speak the Model Context Protocol: Claude Code, Claude Desktop, and any MCP-compatible client.

# One-time setup
pip install httpx fastmcp

# Then in your Claude Code config (or Claude Desktop config):
#   "vx": { "command": "python", "args": ["/path/to/vx_mcp_server.py"] }

See mcp/INSTALL.md for the full Claude Desktop / Claude Code JSON snippets and api/examples/claude_mcp_recipes.md for seven worked agent recipes (BRCA1 figure, coverage walk, GC analysis, subsampled-BAM inspection, bookmark tour, SVG export, video recording).

---

Drive VX from a script (HTTP API)

Every MCP tool is also a plain HTTP endpoint. The full schema is api/openapi.yaml (OpenAPI 3.1, validated). Three runnable examples:

• bash + curl: api/examples/curl_navigate.sh
• Python (stdlib only): api/examples/python_client.py
• Claude agent recipes: api/examples/claude_mcp_recipes.md

# Navigate to BRCA1
curl -X POST -H 'Content-Type: application/json' \
     -d '{"command":"navigate","params":{"chromosome":"17","start":43044295,"end":43125483}}' \
     http://127.0.0.1:9876/command

# Screenshot the viewport at 2x scale
curl -s 'http://127.0.0.1:9876/screenshot/viewport?scale=2.0' \
     -o brca1.png

# Run a peak-calling analysis on the visible region
curl -X POST -H 'Content-Type: application/json' \
     -d '{"command":"run_analysis","params":{"analysis":"SimplePeakCalling","tracks":["ChIP_signal"]}}' \
     http://127.0.0.1:9876/command

---

What you get out of the box

Built-in computational analyses (50+, all in background thread)

Signal & quantification: SignalDifference, SignalRatio, ZScoreNormalization, RPKM, TPM, CPM, PerExonCoverage, BinStatistics, RegionDifferential

Peak calling & enrichment: SimplePeakCalling, EnrichmentOverInput, FeatureProminence, CoverageAnomalies, VariantDensityHotspots, GapDetection

Sequence: GCContent, CpGDensity, MotifSearch, RepeatDensity, NucleotideFrequency

Intervals: IntervalIntersection, IntervalSubtraction, IntervalUnion, ClosestFeature, FeatureCounts

Variants: VariantAnnotation, TsTvRatio, VariantDensity, AlleleFrequencySpectrum, VariantIntersection, CodingImpact

Alignment QC: MappingQualityDistribution, InsertSizeDistribution, StrandBias, DuplicateRate, SoftClipFrequency, ChimericReadDensity, ReadLengthDistribution, PerBaseMismatchRate

Hi-C / interaction: ContactFrequencyVsDistance, InteractionEnrichment, TADBoundaryDetection

Cross-track: SignalCorrelation, SignalAtFeatures, HeatmapMatrix, ProfilePlotData

Each analysis writes BED / BedGraph / TSV / CSV output to a configurable results directory and can optionally create a new track from the result.

File formats supported

Sequence: FASTA, FASTA.gz, indexed FASTA (.fai). Annotation: GTF, GFF3, GFF, GFF.gz. Alignments: BAM (with .bai index), CRAM (limited). Signal: BigWig, BedGraph. Intervals: BED, BED.gz, BigBed. Variants: VCF, VCF.gz (with .tbi index). Interaction: Hi-C contact matrices (.cool, .mcool, planned).

See docs/FILE_FORMATS.md for the full matrix.

Export formats

Static: PNG (full window, viewport, single-track), SVG (vector, viewport, single-track), BED / BedGraph / TSV / CSV (analysis output), session save / restore. Animated: MP4 (H.264), WebM (VP9), GIF direct viewport capture via ffmpeg pipe.

---

Known limitations in v0.9.0

These are documented up front so you can decide whether v0.9.0 fits your workflow.

• Windows ZIP ships a console-subsystem build, launching from Explorer opens a cmd window alongside the GUI. Cosmetic only; the GUI itself works identically. A no-console (GUI subsystem) build is planned for v0.9.x.
• Windows .exe installer is not included in v0.9.0 (Inno Setup workflow needs further work on the build host). ZIP-only Windows distribution for this release.
• macOS bundle is ad-hoc signed only (no Developer ID, no notarisation). First launch requires right-click → Open to clear the one-time Gatekeeper confirmation. The HTTP API and the relocatable bundle pass smoke tests; full GUI verification on physical Apple Silicon is pending (current testing has been on a software-rendered VM tier).
• First-region BAM fetch on cold ≥20 GB BAMs can take ~2 min while the OS warms the BAM index page cache. The load-status strip surfaces this to the user; subsequent regions are fast.
• No web/cloud version. VX is a desktop application. If you need an embeddable browser-tab viewer, JBrowse 2 may suit you better.

See docs/TROUBLESHOOTING.md for runtime troubleshooting.

---

Repository layout

This repository is the public-facing distribution tree for VX. The D source code that produces the binaries is held in a separate private development tree under a proprietary licence (see Source code below).

Folder	Contents	Licence
mcp/	Open-source Python MCP server for Claude / agents	MIT
api/	OpenAPI 3.1 HTTP API spec + curl/Python/Claude examples	MIT
docs/	User guide, file-format guide, keyboard shortcuts, troubleshooting, build provenance, screenshots	CC-BY-4.0
binaries/	Release-asset notes + SHA-256 verification recipes	CC-BY-4.0
Pre-compiled binaries	Attached to each tagged release under Releases	CC-BY-NC-ND-4.0

The per-folder LICENSE matrix is reflected by LICENSE-binaries.txt, LICENSE-docs.txt, mcp/LICENSE, and api/LICENSE.

---

Source code

The VX D source code is proprietary and confidential and is held under a separate licence by Biocodecs Group / Arnaroo Ribologicals. It is not distributed via this repository.

The pre-compiled binaries are licensed under CC-BY-NC-ND-4.0 (non-commercial, no-derivatives, with attribution). For commercial licensing, OEM integration, white-label deployment, or source-code access: contact@biocodecs.org.

The MCP server (mcp/) and the HTTP API specification (api/) ARE open source under MIT and can be reused freely.

---

Citing

If you use VX in research, please cite the accompanying manuscript and the Zenodo DOI of the specific release you used.

Cite the project (all versions, always resolves to the latest release):

Shirokikh, N. E., & Cleynen, A. VX: an AI-agent-controllable desktop genome and transcriptome browser with a programmable analysis framework. Zenodo https://doi.org/10.5281/zenodo.20234450

Cite bioRxiv preprint:

VX: an AI-enabled desktop genome viewer and transcriptome browser with a programmable analysis framework Nikolay E Shirokikh, Alice Cleynen bioRxiv 2026.05.17.725790; doi: https://doi.org/10.64898/2026.05.17.725790

Cite a specific release (v0.9.0.1, snapshot of the v0.9.0 "Aardvark" public pre-release):

Shirokikh, N. E., & Cleynen, A. (2026). Arnaroo/VX: VX 0.9.0.1 — archival re-publish. Zenodo. https://doi.org/10.5281/zenodo.20234451

BibTeX:

@software{vx_genome_viewer,
  author       = {Shirokikh, Nikolay E. and Cleynen, Alice},
  title        = {{VX: an AI-agent-controllable desktop genome
                   and transcriptome browser with a programmable
                   analysis framework}},
  year         = {2026},
  publisher    = {Zenodo},
  doi          = {10.5281/zenodo.20234450},
  url          = {https://doi.org/10.5281/zenodo.20234450}
}

Once the accompanying manuscript is published, this block will gain an additional reference to the journal article; the Zenodo DOIs above will continue to resolve to the archived software snapshots.

---

Build provenance

For full transparency on how each release binary was produced, including compiler version, flags, LTO settings, bundled libraries, signing flow, recommended downstream verification, see docs/BUILD_PROVENANCE.md.

---

Keywords

genome browser · transcriptome browser · sequence viewer · read alignment viewer · NGS · next-generation sequencing · RNA-seq · ChIP-seq · ATAC-seq · CLIP-seq · Hi-C · long-read sequencing · Nanopore · PacBio · Illumina · variant viewer · VCF viewer · BAM viewer · GTF viewer · GFF viewer · FASTA viewer · BigWig · BigBed · BedGraph · OpenGL · GPU-accelerated visualisation · GTK3 · D language · LDC · cross-platform bioinformatics · interactive genome visualisation · genome visualization · transcriptome visualization · MCP · Model Context Protocol · Claude · Anthropic · AI agent · LLM-driven analysis · agentic bioinformatics · HTTP API · peak calling · GC content · CpG density · motif search · Ts/Tv · variant density · alignment QC · MAPQ · insert size · strand bias · soft clip · TAD boundary · contact-vs-distance · heatmap · profile plot · signal correlation · feature counts · RPKM · TPM · CPM · coverage anomaly · gene structure · exon · intron · UTR · CDS · isoform · transcript · subsampling · async loading · video recording · screenshot · screencast · SVG export · reproducible figures · publication-grade · desktop bioinformatics · IGV alternative · JBrowse alternative · Tablet alternative · SeqMonk alternative · Pileup.js alternative.