A lab information system you run yourself. Sample inventory, sequences, protocols, projects and an audit trail — on your own machine, with the database in a file you can copy.
Most lab software of this kind is a hosted subscription: your data lives on someone else's server, priced per seat, and leaving means an export ticket. GeneNet is the other arrangement. It is a single Next.js application over a SQLite file. If you can copy that file, you have your data.
Status: in use, but young. The schema is still moving and there is no upgrade path between versions yet. Read Before you rely on it.
Inventory — samples with freezer/rack/box/position, freezer contents, collections, and a search across every record type.
Sequences — a viewer with restriction mapping against 570 enzymes from
REBASE, Type IIS support for Golden Gate, feature
tracks, translation, and import from GenBank, FASTA, SnapGene .dna, or an
NCBI/UniProt accession. Plus:
- Alignment — Gotoh affine-gap, used to verify a Sanger read against its reference. Tries both orientations and reports which one fit.
- CRISPR — guide design for SpCas9 and friends, scored with published rules rather than a fitted model, so the panel can show you why a guide lost points. Off-targets are searched within the loaded sequence, not genome-wide.
Protocols and work — SOPs with versions, experiments, projects, tasks, gels, bioreactors, proteins, reports.
Records that hold up — every write is captured with who, what and when. Electronic signatures carry a meaning (authorship, review, approval, responsibility) and are hashed against the content they signed, so a signature stops reading as current the moment the record changes underneath it.
An API — bearer tokens with read or write scope, for instruments and scripts. Writes are attributed to the person the token belongs to, so the audit trail stays honest. Revoking someone's account revokes their tokens.
Node 24 or newer. The test suite uses Node's built-in runner and its native TypeScript support, so there are no test dependencies to install.
git clone https://github.com/bioinfocodex/genenet.git cd genenet npm install cp .env.example .env # then set SESSION_SECRET npx prisma migrate dev npm run dev
Open http://localhost:3000. The first account you create becomes the admin.
npm test # 196 tests npm run typecheck npm run build
Point DATABASE_URL somewhere outside Dropbox, OneDrive, iCloud or Google
Drive. Two machines syncing one SQLite file will corrupt it — this is a property
of file sync, not a GeneNet bug. Startup checks the path and warns you.
To share a database between people, run one instance on one machine and have
everyone point a browser at it. Back it up: GeneNet takes an hourly VACUUM INTO
snapshot and keeps the last 14.
Worth saying plainly, because the alternative is you finding out later:
- No upgrade path yet. Schema changes may need a manual migration.
- Audit trail, not a validated system. The record-keeping is built along 21 CFR Part 11 lines — attribution, tamper-evident signatures, no silent edits. That is a foundation for a validated deployment, not a substitute for one. Validation is something you perform against your SOPs, and no software can hand it to you off the shelf.
- Off-target search is local. CRISPR off-targets are found within the sequence you loaded. For a genome-wide search, use a genome-wide tool.
- Single-server. SQLite in WAL mode, one Next.js process. It suits a lab. It is not built for a multi-tenant deployment.
Issues and pull requests are welcome. npm test && npm run typecheck should
pass; CI runs both plus a build against a real database.
The code aims to explain its reasoning where the reasoning is not obvious —
see src/lib/api-auth.ts or src/lib/crispr.ts for the register. Comments
explain why, not what.
MIT. Use it, change it, ship it, sell it — commercially or not. The only condition is that the copyright notice travels with it.
GeneNet is free, and funded by support and hosting for labs that want it rather than by charging per seat.