Loading...
Molecular visualisation for bioscience
Molecular Simulation and Visualisation for Science and Research
Loading interactive ATP molecule.
The problem
Structural biology still runs on scripts and clusters.
To see a protein the way you mean it, you write PyMOL. To dock a ligand, you provision a box, install Vina, prep the receptor by hand. To screen a library, you babysit a queue. The tools are three decades deep — and three decades hard.
You script what you want to see
A publication figure means PyMOL macros, selection algebra, and a render config — a language, not a sentence.
You babysit the infrastructure
Docking means a box, Vina installed, a receptor prepped by hand, a grid box guessed, a queue watched.
The tools are 30 years deep — and 30 years hard
The capability exists. The friction is the specialist software, the setup, and the wait between you and the answer.
What if you could just say it?
See it work
Watch a sentence become science.
One instruction. The director loads the structure, runs the BLAST, finds the pocket, docks the ligand, and renders the movie — and shows every step it took.
DirectorAzure OpenAI · GPT-5.5
Load 1CRN and colour it by hydrophobicity. Find its closest homologs with BLAST and colour by conservation. Detect pockets, dock ibuprofen into the top one, run a short MD, then render a publication movie.
0/7
moleculestud.io/studio
Illustrative replay. Every step maps to a real director tool. Visualisation and in-browser movies are available today; BLAST, pockets, docking and molecular dynamics are rolling out in pilot as compute lanes come online.
One thread. The whole workflow.
Not a suite of apps to stitch together — a single conversation that carries you from a raw structure to a shareable result.
01
Visualise
Live
Load any PDB or AlphaFold-DB structure, style it, colour by property, measure, annotate, cut away.
02
Search & align
Pilot
BLAST for homologs, colour by per-residue conservation, superpose related structures.
03
Find pockets
Pilot
Detect candidate cavities with fpocket — a hole of the right shape, ranked for you.
04
Dock & screen
Pilot
Dock a ligand or screen a library with AutoDock Vina — a computational ranking, not measured affinity.
05
Simulate
Pilot
Relax and probe dynamics with an OpenMM molecular-dynamics run, proposed as a costed job.
06
Render & share
Live
Turntable movies in the browser, a reproducible MolViewSpec to export, a scene to share by link.
No scripts. Here’s the difference.
Same result. On the left, the way it works today. On the right, the way it works here.
the old way — by hand
# 1. figure, in PyMOL
fetch 1crn, async=0
hide everything; show cartoon
set_color hydro, [0.2,0.6,0.9]
spectrum b, blue_white_red
# 2. homologs, on the CLI
blastp -query 1crn.fasta -db pdbaa \
-outfmt 6 -out hits.tsv
# 3. prep + dock, on a box you manage
prepare_receptor -r 1crn.pdb -o 1crn.pdbqt
fpocket -f 1crn.pdb
vina --receptor 1crn.pdbqt --ligand ibu.pdbqt \
--center_x 8 --center_y 6 --center_z 12 \
--size_x 20 --size_y 20 --size_z 20 \
--exhaustiveness 16the Molecule Studio way
“Load 1CRN, colour by hydrophobicity, BLAST for homologs, find pockets, and dock ibuprofen into the top one.”
One thread. No installs, no receptor prep, no grid-box guessing, no cluster. The director does the work and shows you exactly what it ran.
Reproducible by design
It shows its work.
An AI you can trust with your science is one you can check. Every result is an artifact you can inspect, export, re-run, and share.
Every action is a MolViewSpec
The scene the director builds is a portable spec you can export, re-run, and diff — not a screenshot you can’t reproduce.
Every job is a record
Inputs, parameters, and versioned outputs are captured per run, so a docking result can be inspected and re-run, not just trusted.
Share a scene by a link
Send a reviewer the exact structure, colouring, and camera — reproducible on their screen, no install required.
Built for the way you actually work.
From a first look at a fold to ranking a ligand library — the same chat, different intent.
Live
Figures & structure gazing
Publication-ready views and turntable movies from a sentence, in the browser.
Pilot
Homology & conservation
BLAST for relatives and paint conservation onto the fold to reason about function.
Pilot
Docking & virtual screening
Pose a ligand or rank a library against a detected pocket — ranking, not proof of binding.
Pilot
Molecular dynamics
Relax a complex and probe its motion with OpenMM, proposed and costed before it runs.
Programmable, not just a website.
The same director that drives the app is a versioned REST API and an MCP server. Call it from your pipeline, a notebook, or an AI agent: natural language in; structured ViewerOps, costed job proposals, and results streamed back.
describe → drive → submit → results
curl https://moleculestud.io/api/v1/molecular/director \
-H "Authorization: Bearer $MOLECULE_STUDIO_API_KEY" \
-H "Content-Type: application/json" \
-d '{
"prompt": "Load 1CRN, find pockets, dock ibuprofen into the top one",
"headless": true
}'
# Streams the director's ViewerOps, job proposals, and results over SSE.Security & compliance
Your structures and results, handled with care.
The same data-protection posture we’d want as scientists: encrypted, access-scoped, and independently audited.
GDPR-compliantEncrypted in transit & at restSOC 2 Type II — audit in progress
Supported by
Start PAYG. Upgrade when your usage becomes predictable.
Use one credit balance for AI direction, molecular rendering, inference, and simulation. Every expensive action is estimated before it runs.
Pay As You Go
No monthly feebuy credit only when you need it
Estimate every paid action before it runs and spend only on the work you use.
- £5 free credit on signup, plus £5 more when you add a card
- AI direction and cloud rendering
- Top-up bonuses: 10% at £200 and 20% at £500
- Standard job queue
- Full REST API and MCP access
Most popular
Pro
£49per month, excluding VAT
A predictable monthly allowance with better rates when you need more.
- £60 of monthly credit
- 10% off additional usage
- Priority job queue
- Higher API and MCP limits
- Priority support
Enterprise
Customscoped with your organisation
Negotiated terms for organisations with security, procurement, or deployment requirements.
- Everything in Pro
- Negotiated usage and overage
- Data residency and custom retention
- On-premises or private-cloud deployment
- SLA and security review
FAQs
What it does, what’s live vs pilot, how we keep the science honest, cost, data, and automation — answered before you run a thing.
Stop scripting. Start describing.
Open Molecule Studio and drive real structural biology from a chat — no PyMOL, no cluster, no setup.