Tool-review

A five-person molecular biology lab easily spends $6,000 a year on per-seat subscriptions for Prism, SnapGene, FlowJo, and BioRender. I tested the free, open-source replacements on real bench data — from 4PL IC50 curves and diagnostic digests to FCS gating — and looked honestly at where the free stack wins and where it still hurts.

A free, MIT-licensed web app that treats your cloning history — not just the final map — as the thing worth saving. I ran it self-hosted, cloned URA3 into pUC19 from a Python script, and looked hard at where it falls short of SnapGene and Benchling.

A Harvard structural biology lab just open-sourced the software it runs its own bench on — notebook, registry, cryo-EM grid tracking, and an AI agent, all self-hosted under AGPL. I installed it and ran it. Here is what actually works, and what you should know before trusting it with your lab records.

You installed samtools and it broke your entire Python setup. Or DESeq2 needs R 4.3 but your system has R 4.1. Here is how Conda solves the dependency nightmare that every biologist runs into — with real commands, real installs, and real output.

Choosing a guide RNA that cuts your target is only half the battle. You also need one that doesn’t cut anywhere else. This post breaks down the MIT and CFD scoring systems, how off-target algorithms actually work, and which tools are best for predicting Cas9 and Cas12a off-targets.