What you can do
Ask scientific questions
Investigate genes, variants, proteins, pathways, disease mechanisms, targets, compounds, and experimental results with traceable sources.
Run bioinformatics
Analyze transcriptomic, genomic, proteomic, multi-omic, and tabular research data with reproducible scripts, tables, and figures.
Conduct Deep Research
Build a directed literature review, inspect source decisions and evidence gaps, and produce a fully cited PRISMA-style paper.
Plan complex work
Review the approach, assumptions, constraints, validation, deliverables, and success criteria before execution begins.
Use biological models
Work with protein structures, molecular viewers, protein design tools, embeddings, and other specialized biological capabilities.
Write and review papers
Create editable scientific documents, work with citations, and continue drafting with AI-assisted writing tools.
How work is organized
- Sessions keep a request, its follow-up messages, reasoning, and generated outputs together.
- Projects group related sessions, files, papers, and jobs around a study.
- Files contain uploads and outputs produced during research or analysis.
- Compute shows active and completed execution, including background research and analysis tasks.
- Write Paper provides an editable workspace for manuscripts, reports, and other scientific documents.
Explore the docs
Quickstart
Start a session, add files, choose a workflow, and inspect the result.
Platform overview
Understand the reasoning, knowledge, compute, and output layers.
Plan Mode
Approve a detailed plan before Purna performs complex work.
Deep Research
Run transparent, source-backed literature reviews.
Bioinformatics
Analyze biological datasets and inspect reproducible outputs.
Usage and billing
Understand queries, credits, runtime, storage, and plan limits.
Get started
Open Purna, sign in, and follow the Quickstart guide.Purna supports research and decision-making, but it does not replace expert scientific, clinical, regulatory, or experimental review. Verify important claims and outputs before using them in high-stakes work.