Accelerated virtual screening,
at production speed.

MolDyn — Accelerated Virtual Screening

7 targets validated on DUD-E. EF@1% up to 12.92×. Cloud SaaS — no infrastructure needed.
16.77 ms core search over 2.6M molecules · ~200 ms end-to-end via HTTPS.
Screen 2.6M compounds for the cost of your screening team.

What MolDyn is for

Built for protease and hydrolase programs.

Only the targets listed below have been validated on DUD-E (2026-04-18). For anything else, we run a mini-benchmark on your specific target before committing to a paid engagement.

🦠

Antiviral

HIV-1 protease — EF@1% = 12.92
SARS-CoV-2 Mpro (PDB 6LU7) — EF@1% = 50 (case study)

🧬

Aspartic proteases

Renin (hypertension) — EF@1% = 7.36
BACE-1 (Alzheimer's) — EF@1% = 4.36

⚖

Classical serine proteases

Thrombin — EF@1% = 6.62
Trypsin (broad proteolysis) — EF@1% = 5.60

🔬

Hydrolases

Acetylcholinesterase (neuro) — EF@1% = 8.93
Glucocerebrosidase (Gaucher disease) — EF@1% = 9.89

Target not listed? [email protected] — we will run a mini-benchmark before you commit.

The problem

Virtual screening bottlenecks drug discovery.

Hit discovery campaigns stall at the library-screening step. Modern compound libraries have outgrown the tools built to search them, and every month lost in screening is a month off your program timeline.

Slow

Legacy screening infrastructure takes hours or days per target. Teams triage around the tool, not the biology.

Expensive

Commercial docking suites price per seat and per CPU-hour. A single virtual screen can cost tens of thousands of dollars in compute alone.

Linear scaling

Every new target, every library refresh, every mode explored — screen time grows with library size. There is no economy of scale.

The math

The math is simple.

Same budget. Millions of compounds screened.

👤

Traditional approach

1 senior medicinal chemist

$250,000–$350,000 / year (salary + overhead)
100–1,000 compounds screened per week
Results in weeks to months
Expertise required for each target class
Linear scaling — more targets = more headcount

Same price

⚡

MolDyn Starter Plan

Hosted HTTPS API

$300,000 / year ($25K / month)
2,600,000 compounds screened per query
Results in minutes
Works across 7 validated target classes
Same speed for 1 target or 10

MolDyn is a first-pass screening tool, not a replacement for medicinal chemistry expertise. Your chemists focus on hit-to-lead optimization while MolDyn handles the initial screening funnel.

The solution

MolDyn by CellsWave.

A production-grade proprietary screening platform built for multi-million compound libraries, delivered as a hosted HTTPS API.

16.77 ms

Core search time

Sub-20-millisecond search across 2.6 M ChEMBL compounds on a single GPU. End-to-end HTTP latency including SMILES encode, auth, and network is ~103 ms per call.

12.92×

EF@1% on HIV-1 protease

Validated against the DUD-E gold-standard benchmark. Seven targets screening-ready (EF@1% ≥ 4) across aspartic proteases, classical serine proteases, and hydrolases.

SaaS

Hosted HTTPS API

No GPU, no Docker, no install. Submit SMILES over HTTPS, get ranked hits back. Queries are processed in memory and discarded; only a SHA-256 hash is logged.

Scalability

Millisecond screening. Every query.

Screening latency stays flat across queries — the whole library fits in GPU memory and is scored in a single pass. You submit a molecule, you get ranked hits back before most tools have finished parsing their input.

16.77 ms

Core GPU search

Single query against millions of molecules. Results in milliseconds, every time.

~300 ms

End-to-end per query

Full round-trip over HTTPS. Submit molecule, receive ranked hits.

2.6M+

Default library included

A large reference library is pre-indexed and ready to screen on day one. Add your own custom library anytime.

MolDyn is a fast first-pass screening tool. Use it before physics-based docking to prioritize your best candidates.

DUD-E benchmark

Validated against the industry standard.

Every claim on this page comes from the public DUD-E benchmark — the same dataset used to evaluate industry-standard docking and similarity methods.

Validated targets (EF@1% ≥ 4)

Target	Disease area	EF@1%	AUROC
HIVPR	HIV / AIDS	12.92×	0.781
GLCM^*	Gaucher disease	9.89×	0.581
ACES	Alzheimer's / neuro	8.93×	0.618
RENI	Hypertension	7.36×	0.715
THRB	Thrombosis	6.62×	0.744
TRY1	Broad protease	5.60×	0.722
BACE1	Alzheimer's	4.36×	0.711

Outside current envelope (EF@1% < 2)

Target	Disease area	EF@1%	Note
FA7	Anticoagulation	0.90×	Contact us to discuss
FA10	Anticoagulation	0.75×	Contact us to discuss
UROK	Thrombolysis	0.00×	Contact us to discuss

Envelope. Aspartic proteases, classical serine proteases (thrombin / trypsin family), and serine / glycosidase hydrolases are supported. Coagulation-cascade serine proteases (factor Xa, factor VIIa, urokinase) fall outside the current envelope — a known limitation shared with several published ligand-based methods on these targets.

EF@1% is our operational metric — the enrichment of true binders in the top 1% of a screen. EF@1% ≥ 4 means a screening-ready target. AUROC is reported for reference. Protocol: 3-query consensus, leave-N-out, DUD-E decoys (Mysinger et al., J. Med. Chem. 2012).
^* GLCM n = 553 (small sample) — EF@1% stable but with wider confidence interval than larger sets.

Industry-standard dataset Validated on the public DUD-E benchmark — the same dataset used to publish results for industry-standard docking and similarity methods.

Published methodology Mysinger et al., J. Med. Chem. 2012. No private test sets, no cherry-picked targets.

Independent validation Full benchmark report and per-target breakdown available upon NDA. See the public report →

Deployment

How MolDyn works.

Submit a query via HTTPS. All computation runs on CellsWave GPU infrastructure. Results come back in milliseconds.

Cloud SaaS

Cloud SaaS — Hosted by CellsWave

HTTPS API · API-key auth · Usage-metered billing

Submit molecules via HTTPS API. All computation on CellsWave GPU infrastructure. Results in milliseconds.

No infrastructure to manage
API key authentication
Usage-metered billing
SLA-backed uptime

Pricing

Start with a Proof of Concept. Convert when hits are in hand.

Flat-fee 8-week PoC on one real target, monthly plans from Starter through Enterprise, and a royalty-based Partner tier for academic and translational groups. Full breakdown on the pricing page.

Proof of Concept

$75,000

one-time, 8 weeks

1 protease target from your pipeline
Ranked hit list + reproducibility pack
100% credit toward a monthly plan if you convert within 60 days

Frequently asked questions.

How is my data protected?

Your SMILES are transmitted over HTTPS, processed in memory on CellsWave servers, and immediately discarded after your query. We log only a non-reversible SHA-256 hash for abuse detection. No compound data is retained after query completion. All usage is covered by our DPA and NDA.

What molecule formats do you support?

SMILES strings and SDF files for ligands; standard PDB format for protein structures. PDB IDs are auto-downloaded from the RCSB if not in your local cache.

Which target classes are supported?

MolDyn v1.0 is validated and optimized for protease and hydrolase targets. Contact us at [email protected] to discuss your specific target.

How does MolDyn protect my proprietary molecules?

Queries travel to CellsWave over TLS 1.3, are processed in memory, and are discarded after the response is returned. Nothing is written to disk beyond a SHA-256 hash of the query string used for abuse detection — the hash is non-reversible and cannot be mapped back to a SMILES. Ranked results are returned to your client and are not retained by CellsWave. Enterprise engagements sign a mutual NDA and a GDPR Article 28 DPA before any query traffic begins.

What is the difference between trial and paid?

Trial gives 1,000 API calls for 30 days so you can reproduce the benchmark and screen your real target against the full library. Paid plans remove the call limit and add priority support.

Accelerated virtual screening,at production speed.