Biologists learn to keep lab notebooks in Bio 101. They document experimental designs, observations, what worked, what didn’t.

But somehow, this fundamental practice gets lost when we move to computational biology.

I’ve met countless bioinformaticians and data scientists who can tell you the exact pH of their last buffer, but can’t remember why they chose specific parameters for an analysis they ran last month.

Here’s why I think every computational biologist should keep a lab notebook:

➡️ SCENARIO: You run a complex command line tool with 15 parameters. Six months later, you need to reproduce the analysis on a similar dataset.

➡️ WITHOUT A NOTEBOOK: You’re digging through 10,000 lines of bash history, trying to remember if you used –min-coverage 10 or 20, and WHY you made that choice.

➡️ WITH A NOTEBOOK: “Tried –min-coverage 10 initially but got too much noise in low-quality regions. Switched to 20 based on Smith et al. 2023 recommendation for similar tissue type.”

The magic isn’t just recording WHAT you did—it’s capturing WHY you did it. When you document your rationale in real-time, you’re not just helping future you. You’re building institutional knowledge that can be shared, reviewed, and improved upon.

Your notebook becomes a roadmap for scaling analyses, training team members, and catching edge cases before they become problems.

We wouldn’t accept a wetlab scientist who couldn’t reproduce their experiments. Why do we accept computational work that can’t be reproduced?

The best part? Your “lab notebook” can be as simple as a markdown file alongside your code. No fancy tools required. (although I personally am a Confluence fan girl 🤓 )

Do you keep a computational lab notebook? What’s your system for documenting analysis decisions?