Start analyzing bioinformatics data with Omics-OS Cloud — sign up, explore the cockpit, and run your first analysis

Getting Started with Omics-OS Cloud

Get from sign-up to your first analysis in under 5 minutes.

Quick Start

Go to app.omics-os.com and create an account. MFA (TOTP) is required for all accounts — you will set up two-factor authentication during registration.

Sign-in options: email/password, Google, or Microsoft.

Explore the Cockpit

After sign-in you land on the Cockpit — a launcher for all Omics-OS products. From here you can:

Open the Workspace (your analysis environment)
Browse Use Cases (curated analysis workflows you can try instantly)
Open recent sessions
Navigate to Playground, Editor, or Docs. BioVault is shown only as a coming-soon early-access product.

Open the Workspace

Click Workspace in the dock or open a use case. The workspace is a canvas-based environment with chat, data panels, and the OmniBar.

Your first session is created automatically when you send your first message.

Start analyzing

Type a question or request in the chat composer:

You: Search PubMed for recent papers on single-cell RNA-seq of pancreatic beta cells

The supervisor routes your request to the Research Agent, which searches PubMed and returns results in real-time.

The Workspace

The workspace is where analysis happens. It has three areas:

Area	What It Does
Chat panel	Conversation with Lobster. Describe analyses in natural language.
Canvas	Visual workspace. Search results, plots, structures, and pathways appear as interactive nodes.
Data panel	Your files, loaded datasets, plots, team packages, and templates.

The OmniBar (Cmd+Space) overlays the canvas — search databases, ask quick questions, query tables, and capture findings without leaving context. Learn more about OmniBar.

Uploading Data

Drag files into the chat or use the upload button in the data panel.

Supported formats:

Format	Extension	Use Case
AnnData	`.h5ad`	Single-cell data
MuData	`.h5mu`	Multi-modal data
HDF5	`.h5`	Generic HDF5 matrices
CSV/TSV	`.csv`, `.tsv`	Count matrices, metadata
Excel	`.xlsx`	Sample sheets
Loom	`.loom`	Single-cell data
10X MTX	`.mtx`	Sparse matrices from Cell Ranger
Zarr	`.zarr`	Chunked array data

File size limit: 2 GB per dataset on Free tier. Cloud and Enterprise tiers support larger uploads.

After uploading, ask Lobster to load and summarize the data:

You: Load the uploaded file and show me a summary

[Data Expert]
Loading counts.h5ad...
Detected: AnnData (single-cell)
- Cells: 24,531
- Genes: 22,456
- Obs columns: sample_id, condition, batch

Ready for analysis.

Your First Analysis

Example prompts to try:

Literature Search

Search PubMed for recent papers on CRISPR gene editing in bacteria

Dataset Discovery

Find GEO datasets for breast cancer single-cell RNA-seq

Database Search (via OmniBar)

Press Cmd+Space, select the Protein tag, type TP53. A 3D structure node appears on the canvas.

Data Analysis

Load my count matrix, run QC, filter low-quality cells, cluster, and annotate cell types

Multi-Step Workflows

Lobster maintains context within a session. Chain requests naturally:

You: Search for papers on scRNA-seq of liver cells

[Research Agent → 5 papers returned]

You: The third paper mentions a GEO dataset. Find and download it.

[Research Agent + Data Expert → dataset loaded]

You: Run QC and cluster the cells

[Transcriptomics Expert → QC report + UMAP plot on canvas]

Sessions and Persistence

Your work persists automatically:

Files are stored in your workspace and available across sessions
Plots and datasets are tracked per session
Chat history maintains full context
Canvas state (nodes, layout) is restored on reload

To start a new session, click New Session in the workspace. Previous sessions remain accessible from the cockpit or session switcher.

The cockpit shows curated use cases — pre-built analysis workflows you can open with one click. Each use case bootstraps a full session with sample data, chat history, and canvas layout so you can see what a complete analysis looks like before running your own.

Getting Help

Be specific — "Perform DE analysis comparing treatment vs control" works better than "analyze my data"
Provide context — "I have bulk RNA-seq count data from mouse liver, 3 treatment + 3 control"
Use OmniBar — Press Cmd+Space for quick database lookups without interrupting chat
Slash commands — Type /help in chat for available commands

Next Steps

OmniBar

Search databases, query tables, and navigate the canvas

Data Formats

Supported file types and size limits

Example Workflows

Complete analysis examples with real data

Getting Started

Next Steps

On this page