The scientific tools behind real work.
Move from literature and protocol design to molecular modeling, optics, chips, circuits, and fluid dynamics without leaving the same inspectable research workspace.
scientific tools, packages & resources
One searchable catalog spanning agent operations, managed modeling software, and the scientific foundations underneath them.
Select a category
What kind of scientific work do you want to do?
Choose a category below. Its tools, workflows, managed packages, and scientific foundations will appear immediately underneath.
Computational Modeling & Simulation
Choose a scientific field to inspect its managed packages, supported workflows, and application example.
Research, Chemistry, Computer & Foundations
Agent capabilities, the computer and notebooks every plan includes, and the scientific resources underneath them.
Computational Modeling & Simulation
CFD & Aerodynamics
Geometry repair, body-fitted meshing, steady and transient flow, and aerodynamic verification.
-
OpenFOAM
13
Primary CFD environment
Builds and checks volume meshes, solves steady and transient flow, and extracts pressure, velocity, and aerodynamic forces.
Official project -
snappyHexMesh
OpenFOAM 13
Body-fitted meshing
Turns validated surface geometry into a three-dimensional computational mesh with local surface and wake refinement.
Official project -
trimesh
4.12
Geometry audit and repair
Inspects connected bodies, boundaries, normals, scale, triangle quality, and surface deviations before CFD meshing.
Official project -
SU2
8.5
Aerodynamic verification
Provides an independent open-source route for airfoil, external-flow, multiphysics, and design-optimization studies.
Official project
Computational Modeling & Simulation
Optical Simulation
Photon transport, electromagnetic fields, photonic modes, rendering, optimization, and sensitivity analysis.
-
MCX / pmcx
pmcx 0.7.1
GPU photon transport
Traces photon histories through voxelized media for tissue optics, fluence, absorption, and detector-placement studies.
Official project -
Meep
1.34
FDTD electromagnetics
Solves time-domain Maxwell equations for photonic structures, transmission, resonances, fields, and device response.
Official project -
MPB
Photonic band structures
Computes electromagnetic modes and dispersion relations for periodic dielectric structures and photonic crystals.
Official project -
Mitsuba
3.9
Spectral rendering
Models light transport through surfaces, materials, lenses, filters, cameras, and differentiable imaging systems.
Official project -
pymoo
0.6.2
Multi-objective optimization
Explores optical design tradeoffs such as signal, absorption, geometry, robustness, and competing objectives.
Official project -
SALib
1.5.2
Sensitivity analysis
Quantifies how uncertain optical properties and design variables influence simulated outputs and decisions.
Official project
Computational Modeling & Simulation
Semiconductor Design
RTL simulation, lint, synthesis, public-PDK physical design, layout inspection, and device physics.
-
Icarus Verilog
12
RTL simulation
Compiles Verilog and SystemVerilog testbenches, executes them with vvp, and produces inspectable VCD waveforms.
Official project -
Verilator
5.032
SystemVerilog lint
Finds syntax, width, signedness, unreachable logic, and other RTL problems before synthesis or implementation.
Official project -
Yosys
0.52
Logic synthesis
Transforms RTL into technology-independent netlists and reports hierarchy, cells, memories, and structural checks.
Official project -
OpenROAD
public 26Q3 workbench
RTL-to-GDS physical design
Carries suitable designs through floorplanning, placement, clock-tree synthesis, routing, timing, and public-PDK implementation.
Official project -
KLayout
Layout and GDS inspection
Generates and inspects GDS layouts and supports physical-verification work around public semiconductor platforms.
Official project -
DEVSIM
2.10
Semiconductor-device simulation
Solves drift-diffusion and related device equations for diodes, transistors, materials, contacts, and electrical characteristics.
Official project
Computational Modeling & Simulation
Circuit Design
Operating-point, DC, AC, transient, noise, measurement, and device-to-circuit analysis.
-
ngspice
45.2
Circuit simulation
Runs operating-point, DC, AC, transient, noise, Fourier, and measurement-driven analyses for analog and mixed-signal circuits.
Official project -
DEVSIM
2.10
Device-to-circuit modeling
Connects semiconductor-device physics to electrical characteristics that can inform compact models and circuit studies.
Official project -
NumPy / SciPy
Numerical verification
Checks equations, parses solver outputs, compares expected and simulated behavior, and supports parameter and sensitivity studies.
Official project -
JupyterLab
Inspectable analysis
Keeps circuit inputs, equations, waveforms, tables, plots, limitations, and follow-up calculations in a reusable notebook.
Official project
Computational Modeling & Simulation
Computational Chemistry
Molecular properties, conformer search, geometry optimization, electronic structure, and reaction-path calculations.
-
Rowan
Managed molecular workflow layer
Provides the managed API, data model, compute routing, and reproducible workflow layer that connects Vicena to molecular calculations.
Official project -
AIMNet2 / fairchem
Neural interatomic potentials
Provides fast learned energies and forces, including AIMNet2 for organic molecules and OMol25-family models through fairchem.
Workflow documentation -
xTB / CREST
Semiempirical chemistry and conformers
Runs GFN-family semiempirical calculations and supports efficient geometry, energy, descriptor, and conformer-search workflows.
Official project -
PySCF / GPU4PySCF
Molecular electronic structure
Runs Hartree–Fock and density-functional theory, with GPU-accelerated presets for routine and careful quantum-chemistry work.
Official project -
Psi4
Wavefunction quantum chemistry
Provides a modular quantum-chemistry engine for Hartree–Fock, density-functional, and correlated electronic-structure methods.
Official project -
Quantum ESPRESSO
Periodic plane-wave DFT
Handles periodic materials, plane-wave electronic structure, crystal geometries, and related solid-state calculations.
Official project -
geomeTRIC / Sella / ASE
Geometry optimization
Coordinates molecular and periodic geometry optimization, constraints, transition-state work, and crystal-cell relaxation.
Workflow documentation -
RDKit
Molecular informatics and preparation
Handles structures, stereochemistry, conformers, descriptors, fingerprints, hydrogens, bond orders, and chemistry-aware analysis.
Official project
Computational Modeling & Simulation
Molecular Dynamics
Protein and protein–ligand dynamics, pose stability, trajectory analysis, and supported relative binding free-energy workflows.
-
Rowan
Managed molecular workflow layer
Provides the managed API, data model, compute routing, and reproducible workflow layer that connects Vicena to molecular calculations.
Official project -
OpenMM / OpenFF
GPU molecular dynamics
Builds explicit-solvent protein–ligand systems with OpenFF ligand parameters and runs GPU-accelerated dynamics in OpenMM.
Workflow documentation -
TMD
Relative binding free energy
Runs GPU-optimized free-energy perturbation across ligand transformation graphs for relative binding-affinity prediction.
Workflow documentation
Computational Modeling & Simulation
Docking & Binding
Prepare ligands, explore binding poses, screen candidates, and compare supported affinity predictions and free-energy calculations.
-
Rowan
Managed molecular workflow layer
Provides the managed API, data model, compute routing, and reproducible workflow layer that connects Vicena to molecular calculations.
Official project -
Boltz
Structure and affinity models
Supports Boltz-1, Boltz-2, and Boltz-2.1 cofolding workflows with structured confidence, pose, and affinity outputs.
Official project -
Vina / QVina2 / Gnina
Protein–ligand docking
Searches binding poses with Vina-family or Gnina settings, then supports refinement, deduplication, and pose-quality checks.
Workflow documentation -
TMD
Relative binding free energy
Runs GPU-optimized free-energy perturbation across ligand transformation graphs for relative binding-affinity prediction.
Workflow documentation -
RDKit
Molecular informatics and preparation
Handles structures, stereochemistry, conformers, descriptors, fingerprints, hydrogens, bond orders, and chemistry-aware analysis.
Official project
Computational Modeling & Simulation
Protein Structure Prediction
Sequence alignment, protein structure prediction, and biomolecular cofolding with confidence and complex-structure outputs.
-
Rowan
Managed molecular workflow layer
Provides the managed API, data model, compute routing, and reproducible workflow layer that connects Vicena to molecular calculations.
Official project -
OpenFold3
Open biomolecular cofolding
Predicts three-dimensional biomolecular complexes from protein, nucleic-acid, ion, and small-molecule inputs.
Official project -
Boltz
Structure and affinity models
Supports Boltz-1, Boltz-2, and Boltz-2.1 cofolding workflows with structured confidence, pose, and affinity outputs.
Official project -
Chai-1r
Biomolecular structure prediction
Provides a complementary cofolding model for protein, nucleic-acid, and ligand assemblies through the Rowan workflow.
Official project -
ColabFold / MMseqs2
Multiple-sequence alignment
Builds private, reproducible protein alignments in Boltz, Chai, and ColabFold formats for structure-prediction workflows.
Workflow documentation
3 capabilities
Computer & Notebooks
Every Vicena account, Free included, has its own persistent Linux computer with JupyterLab. The agent runs commands, edits files and works in notebooks that you can open and edit too.
💻 Computer
Commands, files and images on your own Linux computer
The agent runs shell commands on your persistent computer, edits files with patches and looks at the images it produces. Use it to install packages, organize and inspect files, run scripts and do quick calculations.
View methodology and examples
Computer
Commands, files and images on your own Linux computer
The agent runs shell commands on your persistent computer, edits files with patches and looks at the images it produces. Use it to install packages, organize and inspect files, run scripts and do quick calculations.
View methodology and examples
How it works
Each user has one dedicated Linux computer. Files in its home directory and the packages you install persist between conversations, and the files appear in the Files panel. The computer stops when idle and starts again on the next use.
Input
Plain-language instructions for what to install, inspect, organize, calculate or run. You can also give shell commands directly.
Output
Command output shown in the chat. Files that commands create stay on your computer and open from the Files panel or from file links.
Limitations
The computer has no GPU. Deleting or overwriting a file is permanent.
Tips
Use the computer for package installs, file management, scripts and short computations. Ask for a notebook when you want the code, plots and tables kept in one visible document.
Example prompts
- › Generate a sine-wave plot from synthetic data and show me the image
- › Calculate the molecular weight and logP of caffeine from SMILES Cn1cnc2c1c(=O)n(C)c(=O)n2C with RDKit
Notebooks
Shared Jupyter notebooks for analysis, simulations and plots
The agent writes and runs cells in a JupyterLab notebook on your computer while you watch. You and the agent edit the same notebook and share one kernel, so the code, outputs and plots stay visible and reusable.
View methodology and examples
Notebooks
Shared Jupyter notebooks for analysis, simulations and plots
The agent writes and runs cells in a JupyterLab notebook on your computer while you watch. You and the agent edit the same notebook and share one kernel, so the code, outputs and plots stay visible and reusable.
View methodology and examples
How it works
JupyterLab runs on your computer. The agent and your open notebook tab share one document and one kernel: cells appear as the agent writes them, output streams in while they run, and variables you define are visible to the agent. It edits cells by their stable IDs, so it does not overwrite your own edits; if you change a cell it was about to edit, it reads the cell again first.
Notebook workflow
Use notebooks for computational work you want to see and keep. Use the computer directly for package installs, file checks and batch scripts.
A visible computational record
The notebook, with its cells and outputs, stays on your computer and appears in Files.
Input
Requests for visible Python work: build a simulation, analyze data, plot results, fix a failing cell, turn a notebook into a report, or continue a notebook you edited.
Output
A live notebook on your computer and, in the chat, each run cell's status and text output. The agent sees plots as images.
Limitations
The computer has no GPU, so GPU training and very large datasets do not fit. Notebook files live on your computer with your other files.
Tips
Say which notebook to continue when several are open.
Example prompts
- › In a notebook, fit this Michaelis-Menten dataset and report Km, Vmax, residuals and a plot: substrate_uM=[5,10,20,50,100,200], rate=[0.08,0.15,0.27,0.50,0.72,0.88].
- › In a notebook, fit a linear calibration curve and report R², the limit of detection and a residual plot: concentration=[0,1,2,5,10], signal=[0.03,0.18,0.35,0.82,1.61].
📚 Scientific Library
Tested guidance and scripts installed on every computer
Vicena installs a read-only library of guidance and Python scripts on every computer. For a task such as reading a PDF, downloading a dataset, predicting a reaction or setting up a simulation, the agent reads the relevant page and runs its scripts in place.
View methodology and examples
Scientific Library
Tested guidance and scripts installed on every computer
Vicena installs a read-only library of guidance and Python scripts on every computer. For a task such as reading a PDF, downloading a dataset, predicting a reaction or setting up a simulation, the agent reads the relevant page and runs its scripts in place.
View methodology and examples
How it works
Scripts declare exact package versions and run with uv, so every user runs the same tested code. Vicena updates the library on every computer start; the agent never copies it into your files.
Output
Script results as JSON or text in the chat, and larger outputs as files in your project folders.
Limitations
Library scripts run on your computer’s CPU. The first run of a script that needs a large model, such as reaction prediction, downloads it first.
Example prompts
- › Read pages 3 to 5 of the PDF I uploaded and extract the table of activation energies
- › Find the dataset behind DOI 10.5281/zenodo.6795835 and download its CSV files
4 capabilities
Research & Discovery
Search the web and public data repositories, read papers and PDFs, and draft or review lab protocols with their sources cited.
Web Search
Search the web and read the sources
Searches the web for papers, patent records, technical specifications, manuals, supplier pages and safety data sheets, and opens the pages it needs before answering. Image search is included.
View methodology and examples
Web Search
Search the web and read the sources
Searches the web for papers, patent records, technical specifications, manuals, supplier pages and safety data sheets, and opens the pages it needs before answering. Image search is included.
View methodology and examples
How it works
The agent searches, opens the most relevant results or the URLs you give it, and reads their text. It follows up with narrower searches when the evidence leaves a gap, and cites the pages it read.
Input
A research question. Include the source types, date range and other constraints that matter.
Output
An answer that cites the pages read, with links, and states what remains unresolved.
Limitations
Only publicly accessible pages can be read; paywalled or unreadable pages leave gaps. Patent search does not establish patentability or freedom to operate.
Example prompts
- › Find recent papers on CRISPR delivery mechanisms in solid tumors
- › Search patents for a biodegradable polymer stent with drug elution
- › Compare specifications of benchtop centrifuges rated for 15,000 RPM
Datasets
Find and download public research data
Searches BioStudies, Zenodo, Figshare and Dryad through their official APIs, turns a DOI, accession or record link into its file list, and downloads the chosen files to your computer.
View methodology and examples
Datasets
Find and download public research data
Searches BioStudies, Zenodo, Figshare and Dryad through their official APIs, turns a DOI, accession or record link into its file list, and downloads the chosen files to your computer.
View methodology and examples
How it works
A library script queries the repositories’ APIs and returns candidate records with titles, DOIs, file counts, sizes and licenses. When several records fit, the agent shows you the candidates instead of guessing. It checks free space before a large download and verifies checksums when the repository provides them. Other repositories are reached through web search and their official download links.
Input
A dataset description, a paper whose data you need, or a DOI, accession or record link.
Output
The files in a folder on your computer, with the record’s title, DOI, link and license.
Limitations
Dryad serves files only with an API token or to a signed-in browser, so the agent may ask you to download them yourself. Closed, embargoed or restricted records need your access. Downloads of more than a few gigabytes need your confirmation.
Example prompts
- › Find a public single-cell RNA-seq atlas of the zebrafish heart
- › Download the data files from https://zenodo.org/records/6795835
📄 PDF Reading
Text, tables and figures from documents on your computer
Reads PDFs on your computer locally: finds the relevant pages, extracts their text and tables, and renders pages as images so the agent can read figures, structures and scanned pages.
View methodology and examples
PDF Reading
Text, tables and figures from documents on your computer
Reads PDFs on your computer locally: finds the relevant pages, extracts their text and tables, and renders pages as images so the agent can read figures, structures and scanned pages.
View methodology and examples
How it works
A library script based on PyMuPDF reports the page count and outline, ranks pages by search terms, and extracts text, tables or page images. The document stays on your computer. The agent cites PDF page numbers and says when it read only part of a document.
Input
A PDF on your computer, such as an uploaded paper, report, datasheet or price list.
Output
Answers with page citations, extracted tables, and page images when figures matter.
Limitations
Tables embedded as images are not detected as tables. Long scanned documents need OCR, which requires installing Tesseract on your computer.
Example prompts
- › Summarize the methods section of the paper I uploaded and cite the pages
- › Extract the price table from page 7 of this catalogue
Protocol Drafting & Review
Evidence-backed protocol drafts and audits
Drafts an operational lab protocol from a scientific objective, or audits a protocol you already have for physical consistency, safety controls and missing details. Both are saved as Markdown files for a scientist to review.
View methodology and examples
Protocol Drafting & Review
Evidence-backed protocol drafts and audits
Drafts an operational lab protocol from a scientific objective, or audits a protocol you already have for physical consistency, safety controls and missing details. Both are saved as Markdown files for a scientist to review.
View methodology and examples
How it works
A draft starts from primary procedural sources found with web search. Every value taken from a source carries its source label; inferred values are listed as assumptions and unknown values as missing details. An audit first rewrites the procedure as structured reagents, vessels and steps, then checks each step against conservation, phase, compatibility and hazard requirements, using the chemistry lookups where they apply.
Input
An objective with the material, organism or method, scale and available equipment; or the protocol text to audit.
Output
A protocol or audit report in Markdown with sources, assumptions, missing details and review findings, starting with the revisions it requires.
Limitations
A draft or audit is not a validated procedure and does not establish that a procedure is safe. A qualified scientist must review it before bench use.
Example prompts
- › Build a protocol for synthesizing gold nanoparticles by citrate reduction
- › Audit this esterification procedure for missing controls and hazards
5 capabilities
Chemistry
Look up compounds and their hazards in PubChem, do exact lab calculations, and run RDKit, thermo and reaction-prediction scripts on your computer.
Compound Lookup & Safety
Identity, properties and hazards from PubChem
Identifies a compound from a name, CAS number, SMILES, InChIKey or PubChem CID and returns its identifiers, formula, masses and computed properties. A safety lookup returns GHS hazards, NFPA ratings, transport class, reactive groups and published incompatibilities, each with its source.
View methodology and examples
Compound Lookup & Safety
Identity, properties and hazards from PubChem
Identifies a compound from a name, CAS number, SMILES, InChIKey or PubChem CID and returns its identifiers, formula, masses and computed properties. A safety lookup returns GHS hazards, NFPA ratings, transport class, reactive groups and published incompatibilities, each with its source.
View methodology and examples
How it works
Vicena queries PubChem live through a rate-limited client. The agent combines the records with other results itself; for example, it compares the hazard statements and incompatibilities of two reagents before advising on mixing them.
Input
A compound name, CAS number, SMILES, InChIKey or CID.
Output
CID, formula, masses, SMILES, InChI, XLogP, TPSA, hydrogen-bond counts, CAS numbers and synonyms; for safety, the GHS signal word, pictograms and hazard statements with sources.
Limitations
Coverage is limited to what PubChem records. A missing hazard statement is not evidence that a compound is safe.
Example prompts
- › What are the GHS hazards of acetonitrile, and what is it incompatible with?
- › Give me the CAS number, formula and TPSA of ibuprofen
Lab Calculations
Exact stoichiometry, solutions, equilibria and kinetics
Molar mass, equation balancing, limiting reactant and yields, solution preparation and dilution, pH of acids, bases and buffers, solubility from Ksp, the ideal gas law, rate laws, Arrhenius parameters and reaction heat with adiabatic temperature rise.
View methodology and examples
Lab Calculations
Exact stoichiometry, solutions, equilibria and kinetics
Molar mass, equation balancing, limiting reactant and yields, solution preparation and dilution, pH of acids, bases and buffers, solubility from Ksp, the ideal gas law, rate laws, Arrhenius parameters and reaction heat with adiabatic temperature rise.
View methodology and examples
How it works
Each calculation is a typed operation on Vicena’s server with units in its field names. It refuses ambiguous input, such as an equation that several independent reactions could balance, instead of guessing. pH comes from an exact charge balance, and reaction heat from Hess’s law with real coefficients and phases.
Input
Formulas, masses or moles, concentrations and volumes, pKa or Ksp values, temperatures, or rate constants.
Output
The computed values with units and the intermediate quantities used.
Limitations
pH calculations assume ideal solutions at 25 °C. Reaction heat needs formation enthalpies from you, the literature or the thermo script.
Example prompts
- › Is 6.2 g of aspirin from 5 g of salicylic acid physically possible?
- › What is the pH of 0.1 M acetic acid with 0.05 M sodium acetate?
- › Estimate the adiabatic temperature rise of neutralizing 1 mol of HCl in 500 mL of water
🔬 Molecule Analysis
Descriptors, similarity and spectra hints with RDKit
Validates SMILES and computes formula, masses, logP, TPSA, hydrogen-bond counts, Lipinski violations and functional groups. Ranks candidates by fingerprint similarity to a query, and lists expected IR bands, NMR shift ranges and MS adducts for a structure.
View methodology and examples
Molecule Analysis
Descriptors, similarity and spectra hints with RDKit
Validates SMILES and computes formula, masses, logP, TPSA, hydrogen-bond counts, Lipinski violations and functional groups. Ranks candidates by fingerprint similarity to a query, and lists expected IR bands, NMR shift ranges and MS adducts for a structure.
View methodology and examples
How it works
Library scripts run RDKit on your computer. Similarity uses Morgan fingerprints and Tanimoto scores. Spectra hints come from a cited table of functional-group ranges, matched with SMARTS patterns.
Input
SMILES strings, or names that the agent first resolves through PubChem.
Output
JSON results that the agent summarizes, with tables for several molecules.
Limitations
Spectra hints are functional-group ranges, not quantum-mechanical predictions. Use a notebook with PySCF for first-principles spectra.
Example prompts
- › Compare aspirin, ibuprofen and naproxen by Lipinski descriptors
- › Which of these ten SMILES is most similar to caffeine?
- › Which functional groups could explain an IR band at 1715 cm⁻¹?
Reaction Prediction
Textbook transformations, products and retrosynthesis
Applies standard textbook transformations, such as esterification, amide formation and hydride reduction, and predicts products or one-step precursors with the ReactionT5v2 model.
View methodology and examples
Reaction Prediction
Textbook transformations, products and retrosynthesis
Applies standard textbook transformations, such as esterification, amide formation and hydride reduction, and predicts products or one-step precursors with the ReactionT5v2 model.
View methodology and examples
How it works
Textbook rules are reaction SMARTS that report every matching site. ReactionT5v2 is a T5 model trained on patent reactions; its published top-1 accuracy is about 97 % for products and 71 % for retrosynthesis on its USPTO benchmarks.
Input
Reactant and reagent SMILES, or a target product for retrosynthesis.
Output
Ranked candidate products or precursors with formulas.
Limitations
Model predictions are hypotheses to check against the literature. They say nothing about yield, selectivity, conditions or safety, and accuracy is lower for chemistry unlike patent reactions.
Example prompts
- › Predict the product of benzoic acid and ethanol under acid catalysis
- › Propose one-step precursors for paracetamol
💧 Solubility & Thermodynamic Data
Measured solubility and tabulated properties
Looks up measured solubility in BigSolDB 2.0, with source DOIs, and returns melting, boiling and critical points, formation enthalpies, vapor pressure, density and heat capacity from the thermo library.
View methodology and examples
Solubility & Thermodynamic Data
Measured solubility and tabulated properties
Looks up measured solubility in BigSolDB 2.0, with source DOIs, and returns melting, boiling and critical points, formation enthalpies, vapor pressure, density and heat capacity from the thermo library.
View methodology and examples
How it works
Library scripts rank solvents for a solute or list a solute–solvent pair over temperature from BigSolDB, and evaluate thermo’s property models at the temperature and pressure you give.
Input
A solute and solvents, or a CAS number with a temperature and pressure.
Output
Measured values with their sources, or property values with units.
Limitations
Solubility is reported only where BigSolDB has measurements; there is no solubility prediction. Thermo properties are limited to compounds in its database.
Example prompts
- › Which solvents dissolve ibuprofen best at 25 °C, according to measured data?
- › What is the vapor pressure of ethanol at 90 °C?
3 capabilities
Molecular Modeling
Keep exact, versioned small-molecule records with the agent, run computational chemistry in notebooks, and submit Rowan workflows for advanced modeling.
⬡ Molecular Workbench
Versioned candidates, proposals, evidence and screening
A shared workbench for exact small-molecule records: named candidates with immutable saved versions, reviewable proposals, RDKit evaluations, evidence, screening collections and next decisions. The panel shows each record as the agent changes it.
View methodology and examples
Molecular Workbench
Versioned candidates, proposals, evidence and screening
A shared workbench for exact small-molecule records: named candidates with immutable saved versions, reviewable proposals, RDKit evaluations, evidence, screening collections and next decisions. The panel shows each record as the agent changes it.
View methodology and examples
How it works
The agent proposes edits instead of changing your draft, and only accepts a proposal when you ask. Evaluations compute RDKit descriptors and an ETKDGv3/MMFF94 conformer for a saved version on your computer. Screening rounds freeze their members, method and criteria.
Input
An objective, target, molecule or uploaded structure. None is required to start research.
Output
Candidate and collection records in the Molecular panel, with versions, evaluations, evidence and decisions.
Limitations
Descriptor screening does not establish solubility, affinity or assay success. Docking and other advanced methods run through Rowan or notebooks.
Example prompts
- › Create a candidate for aspirin and evaluate its descriptors
- › Start a screening round for these five analogues with TPSA below 90 Ų
⚛️ Computational Chemistry
Quantum chemistry and molecular dynamics in notebooks
Runs bounded computational chemistry on your computer, such as conformer searches, PySCF calculations and structure preparation, in notebooks that keep inputs, raw results and plots visible.
View methodology and examples
Computational Chemistry
Quantum chemistry and molecular dynamics in notebooks
Runs bounded computational chemistry on your computer, such as conformer searches, PySCF calculations and structure preparation, in notebooks that keep inputs, raw results and plots visible.
View methodology and examples
How it works
The library’s computational-chemistry pages guide the method choice. The agent uses packages on your computer for lightweight work and suggests Rowan when a task needs advanced remote compute.
Input
A molecule, SMILES, structure file or property you want to compute.
Output
A notebook with raw results, CSV or JSON files, plots, structures and interpretation.
Limitations
The computer has no GPU, so large quantum-chemistry or dynamics runs belong in Rowan.
Example prompts
- › Run a Hartree-Fock calculation on water with PySCF and report the orbital energies
- › Search conformers of aspirin from SMILES CC(=O)Oc1ccccc1C(=O)O and summarize the lowest-energy structures
Rowan Workflows
Advanced molecular modeling on Plus and Ultra
Submits Rowan workflows for pKa, solubility, ADMET, conformers, tautomers, docking, cofolding, molecular dynamics, quantum chemistry and other molecular-modeling tasks, then reads their results.
View methodology and examples
Rowan Workflows
Advanced molecular modeling on Plus and Ultra
Submits Rowan workflows for pKa, solubility, ADMET, conformers, tautomers, docking, cofolding, molecular dynamics, quantum chemistry and other molecular-modeling tasks, then reads their results.
View methodology and examples
How it works
Vicena calls Rowan on your behalf; your computer holds no Rowan credentials. Each submission has a budget in Rowan credits, and Vicena reserves the authorized maximum from your account credits before submitting. The agent asks you to confirm paid submissions.
Input
A molecule, protein or pose and the property or workflow you need.
Output
Workflow results such as predicted properties, poses, trajectories or energies.
Limitations
Rowan needs a Plus or Ultra plan. Each submission reserves credits within its budget cap, and a daily submission limit applies.
Example prompts
- › Predict the pKa values of cetirizine with Rowan
- › Dock this ligand into the prepared target and compare the top poses
1 capabilities
Simulation & Engineering
Design microfluidic chips in the Microfluidics app and run steady flow experiments on them.
⌁ Microfluidics Workbench
Editable chip designs and flow experiments
Designs and edits microfluidic chips in the Microfluidics app from catalog components, runs steady pressure-driven flow experiments, inspects the CAD geometry and exports supplier review packages.
View methodology and examples
Microfluidics Workbench
Editable chip designs and flow experiments
Designs and edits microfluidic chips in the Microfluidics app from catalog components, runs steady pressure-driven flow experiments, inspects the CAD geometry and exports supplier review packages.
View methodology and examples
How it works
Designs are saved as versioned revisions in your account. The network solver computes steady, incompressible channel flow for a saved version, and design checks flag geometry problems before export.
Input
A function the chip should perform, or the geometry and conditions you want to test.
Output
A saved design open in the app, experiment runs with their summaries, and export files on your computer.
Limitations
The solver does not predict transient filling, mixing, valve actuation or optics. A review package or export is not fabrication approval, a quote or an order.
Example prompts
- › Design a glass chip with two inlets joining a 40 mm serpentine channel and open it in the app
- › Run the saved design at 50 kPa inlet pressure and report the flow rates
3 capabilities
Real-world Labs
Find laboratories for physical work, record the evidence for each candidate, and, with your approval, ask the Vicena team to arrange the work.
🧪 Lab Matching
A persistent shortlist with the evidence for each laboratory
Describe a project in plain language. Vicena separates simulation from physical work, turns each experiment into explicit requirements, searches its laboratory directory for candidates, and records the evidence and open questions for each one.
View methodology and examples
Lab Matching
A persistent shortlist with the evidence for each laboratory
Describe a project in plain language. Vicena separates simulation from physical work, turns each experiment into explicit requirements, searches its laboratory directory for candidates, and records the evidence and open questions for each one.
View methodology and examples
How it works
A matching session keeps the work packages and candidates in the conversation. The agent researches the strongest candidates on their official pages and records each requirement as confirmed, probable, unknown, conflicting or a mismatch, with its source. Unknown evidence never counts in a candidate’s favour.
Matching and evidence
Each project keeps a reusable evidence trail in its conversation.
Input
A project objective with any known sample, scale, success criteria, region, timeline, quality standard or budget.
Output
A ranked shortlist with source links, missing confirmations and reasons for rejected candidates.
Limitations
A shortlist does not establish current capacity, pricing, regulatory fit or that a laboratory accepts external work.
Tips
Confirm the sample, scale, controls, acceptance criteria, region, timeline and quality requirements before asking the Vicena team to arrange the work.
Example prompts
- › I need a 100 mg heterocycle reference standard. Plan the experiments and find labs for each step.
- › A polymer part is failing in the field. Plan the analysis and shortlist European labs for each test.
↗ Request Lab Work
Hand a confirmed brief to the Vicena team
After you approve the exact non-confidential brief and contact details, Vicena sends it to its team, who review it, contact laboratories and publish results to your request.
View methodology and examples
Request Lab Work
Hand a confirmed brief to the Vicena team
After you approve the exact non-confidential brief and contact details, Vicena sends it to its team, who review it, contact laboratories and publish results to your request.
View methodology and examples
How it works
The agent shows the brief and asks for your confirmation before submitting. You can follow the request’s status and published results in the same conversation.
Input
Your contact details, a non-confidential brief, the capabilities needed, timing, scale, region, budget and any preferred laboratory.
Output
A request reference with its status and the results the team publishes.
Limitations
A request is not a booking, quote, contract or laboratory acceptance. Do not include confidential structures, unpublished formulas or proprietary protocols.
Example prompts
- › Prepare a non-confidential peptide synthesis request and show it to me before sending
⌕ Request Lab Coverage
Add a missing laboratory, region or capability
Ask the Vicena team to research a named laboratory, website or location that the directory does not yet cover.
View methodology and examples
Request Lab Coverage
Add a missing laboratory, region or capability
Ask the Vicena team to research a named laboratory, website or location that the directory does not yet cover.
View methodology and examples
How it works
With your approval, the agent records the laboratory name, website or location, the relevant capabilities and a non-confidential explanation for the team to review.
Input
Your contact details, a laboratory name, website or location, optional capabilities and a non-confidential summary.
Output
A coverage-request reference for the directory team.
Limitations
This does not contact the laboratory and does not establish that it offers external work.
Example prompts
- › Research analytical testing labs in Finland that are missing from the directory.
Science Stack
31 databases, libraries & modelsThe databases, libraries, and models that Vicena's capabilities build on. Many of the libraries are also installed on your computer for direct use in your notebooks.
PubMed
Databaseby National Library of Medicine (NIH)
The US National Library of Medicine database with over 36 million biomedical citations. The primary source for life sciences and biomedical literature worldwide.
Official projectarXiv
Databaseby Cornell University
Cornell University's open-access repository hosting over 2.4 million preprints in physics, mathematics, computer science, and quantitative biology. Covers cutting-edge research before peer review.
Official projectGoogle Scholar
Databaseby Google
Google's academic search engine indexing the full text of scholarly literature across publishers, disciplines, and formats. Covers papers, theses, books, and conference proceedings.
Official projectOpen web and repository APIs
Databaseby Vicena
Public web sources and official repository APIs used for literature, patent, and dataset discovery and source verification.
PubChem
Databaseby National Center for Biotechnology Information (NIH)
The world's largest open chemistry database, maintained by the NIH. Contains data on 116 million compounds including structures, properties, biological activities, safety information, and patent references.
Official projectBigSolDB 2.0
Databaseby Krasnov et al. (Scientific Data, 2025)
An open dataset of 103,944 experimental solubility values for 1,448 organic solutes in 213 solvents between 243 and 425 K, each with the DOI of the paper it came from.
Official projectRDKit
Libraryby Greg Landrum and contributors
The industry-standard open-source cheminformatics toolkit used by Pfizer, Novartis, and Merck. Handles molecular representation, substructure search, fingerprinting, and property calculation.
Official projectChemPy
Libraryby Bjoern Dahlgren
A Python library for physical chemistry. Solves stoichiometry, balances equations, computes equilibrium constants, and models chemical kinetics from first principles.
Official projectThermo
Libraryby Caleb Bell and contributors
An open-source thermodynamic properties library covering 70,000+ chemicals. Calculates vapor pressure, heat capacity, enthalpy, and phase equilibria using validated correlations from the DIPPR database.
Official projectPySCF
Libraryby Qiming Sun et al.
A quantum chemistry package for Hartree-Fock, DFT, and post-Hartree-Fock calculations. Used in academic research for electronic structure simulations of molecules and materials.
Official projectASE
Libraryby Technical University of Denmark
The Atomic Simulation Environment, a set of tools for setting up, running, and analyzing atomistic simulations. Interfaces with dozens of quantum chemistry and molecular dynamics codes.
Official projectNumPy
Libraryby NumPy community
The fundamental package for numerical computing in Python. Provides N-dimensional arrays, linear algebra, Fourier transforms, and random number generators. The foundation of nearly all scientific Python.
Official projectSciPy
Libraryby SciPy community
Built on NumPy, SciPy adds optimization, integration, interpolation, signal processing, and statistical functions. The go-to library for scientific and engineering computation.
Official projectPandas
Libraryby Wes McKinney and contributors
The standard library for data manipulation in Python. DataFrames make it easy to clean, transform, and analyze tabular data from experiments, simulations, and databases.
Official projectMatplotlib
Libraryby John D. Hunter and contributors
The most widely used plotting library in science. Produces publication-quality figures, histograms, spectra, and scatter plots. Used in thousands of peer-reviewed papers every year.
Official projectPlotly
Libraryby Plotly Technologies Inc.
An interactive visualization library for 3D plots, dashboards, and dynamic charts. Particularly useful for exploring molecular structures, reaction landscapes, and multi-dimensional data.
Official projectPyTorch
Libraryby Meta AI (FAIR)
Meta's open-source deep learning framework, the most popular in academic research. Powers neural networks for reaction prediction, molecular property estimation, and scientific data analysis.
Official projectTransformers
Libraryby Hugging Face
Hugging Face's library providing access to thousands of pre-trained models for NLP, computer vision, and scientific applications. Used for text analysis of papers and chemical language models.
Official projectSymPy
Libraryby SymPy community
A symbolic mathematics library for Python. Solves equations algebraically, computes integrals and derivatives, and simplifies expressions. Useful for deriving analytical solutions to scientific problems.
Official projectScikit-learn
Libraryby INRIA and contributors
The most widely used machine learning library in Python. Provides classification, regression, clustering, and dimensionality reduction algorithms for analyzing scientific datasets.
Official projectOpenCV
Libraryby Intel, Willow Garage, and contributors
The standard computer vision library with tools for image processing, feature detection, and analysis. Used in microscopy, materials characterization, and automated lab image analysis.
Official projectFluids
Libraryby Caleb Bell
A Python library for fluid mechanics calculations. Computes pressure drops, pipe friction factors, and hydraulic properties for chemical engineering and process design.
Official projectJupyterLab
Libraryby Project Jupyter
The open-source interactive development environment used by millions of scientists. Originally developed at UC Berkeley, Jupyter is the standard for reproducible computational research across all scientific disciplines.
Official projectRowan
Libraryby Rowan Scientific
A managed molecular-modeling workflow provider exposed through Vicena for descriptors, quantum chemistry, docking, protein workflows, molecular dynamics, and free-energy calculations.
Official projectOpenMM
Libraryby OpenMM contributors
A high-performance toolkit for molecular simulation used in protein preparation, molecular dynamics, and trajectory-generating workflows.
Official projectOpenFold3
Modelby OpenFold contributors
An open biomolecular structure-modeling route used in supported cofolding and protein-complex workflows when available.
Official projectReactionT5v2
Modelby Tatsuya Sagawa and Ryosuke Kojima (2023)
A T5 language model pre-trained on the Open Reaction Database and fine-tuned on USPTO patent reactions, used for forward product prediction and one-step retrosynthesis.
Official projectMorgan fingerprints
Modelby H. L. Morgan (1965), extended by RDKit
A circular fingerprinting algorithm that encodes the local chemical environment around each atom. Widely used in drug discovery for virtual screening and similarity searching because it captures both topology and atom types.
SMARTS
Standardby Daylight Chemical Information Systems
A pattern language for describing molecular substructures. Used to identify functional groups, pharmacophores, and reactive sites by matching atoms and bonds in molecular graphs.
Arrhenius equation
Modelby Svante Arrhenius (1889)
The foundational model in chemical kinetics describing how reaction rate constants depend on temperature. Developed by Svante Arrhenius in 1889, it remains the standard for predicting reaction speed.
Vicena Computer
Standardby Vicena
Each user’s persistent Linux computer, on every plan. Its home directory holds documents, uploads, protocols, notebooks, datasets and generated results, shown in Vicena’s Files panel.
Ready to try?
Describe your research problem and the agent chooses how to work on it. Every plan, Free included, has its own computer and notebooks; no credit card required.
Try it free