Using AI agents to answer biological questions
The SciLifeLab & Wallenberg National Program for Data-Driven Life Science exists to build exactly the skills this course teaches — and to find the people who will use them.
Interrupt me. This is the one session where asking the obvious question helps everyone.
In 2008 the cost fell off a cliff and never came back. Sequencing stopped being a project and became an assay.
Sequencing output has risen by roughly eight orders of magnitude since the first human genome. The number of people trained to analyse it has not.
384 × 9 × 4 = 13 824 images.
From one plate, in one afternoon. Nobody is going to look at these by eye, and no one is going to count cells by hand.
The domains differ. The workflow does not.
No individual can hold this. The question is not whether we need computation — it is who directs it.
Data-driven work does not replace hypothesis testing. It generates the hypotheses that hypothesis-driven experiments are good at killing.
…and closer to the question.
At CASP14, AlphaFold 2 reached a median GDT_TS of 92.4 — roughly the agreement you would expect between two experimental structures of the same protein.
The 2024 Nobel Prize in Chemistry went to Baker, Hassabis and Jumper. The field has made up its mind about whether this counts.
It aims to go from describing cells to predicting how they change — and then to intervening. It is coordinated a few floors from this lecture.
Programmes at this scale do not fail for lack of models. They fail for lack of people who can turn a biological question into one.
The biologist owns the question; the computational scientist owns the method; and between them sits a misunderstood file format. Most projects do not fail at the model. They fail in translation.
Two large studies, different data, same shape: junior hiring falls, senior employment does not. The mechanism is slower hiring rather than layoffs.
You will not be out-competed by a model. You may be out-competed by someone who directs one better — and unlike the model, that is a skill, which means it can be taught. That is what the next six weeks are.
It answers, it drafts, it explains. Then it stops, and you go and do the work.
Most people's mental model of AI stopped here, in about 2023. The rest of this lecture is about what it missed.
A model this simple has no notion of true. It has a notion of plausible. Hold on to that — it explains every failure later in this lecture.
A chatbot answers. An agent acts, looks at what happened, and tries again.
Nobody typed the fix. Nobody checked whether plate 3 should have been rescaled or excluded, either.
It remembers fine inside one conversation. Nothing survives the compaction and the closed tab except what was written to a file — and it does not matter what kind of file.
Notes, a data dictionary, a script, a README — all of it counts. In tomorrow's lab you start from an empty folder, and what you leave on disk is what survives the week.
A skill is a folder on disk: a SKILL.md saying when to use it and how, plus the scripts it needs. The agent always sees the name and one line; it opens the folder only when the task matches.
Six steps, one afternoon, and you write none of the code. Watch which parts the harness supplies — and which part only you can.
None of this is new. It is the scientific method, pointed at a new instrument.
Every item on the right is domain judgement or communication — neither is taught by a Python tutorial, and both are exactly what the six labs make you practise.
The role below exists to collapse this gap into one person.
Every computer lab rehearses 01 to 05. The final project runs all six, over several weeks, with a real person on the other end.
If you cannot answer all four afterwards, you have not finished the interview — book the second call.
Data owners speak their own language. Assume nothing is defined until you have made it definite.
Ask for one real file before the call ends. Everything you were told will turn out to be slightly different from what is in it.
The agent can do every rung. It cannot tell you that rung two was the wrong quantity to measure.
Everything on the right was already implied on the left. Your job is to make it explicit before the agent has to guess.
Nothing on the last two slides was typing. Every rung of that ladder was a judgement — and three things have to already be in your head before you can make one.
You will translate onto methods you have never run, in fields you do not work in. That is the normal case, not the failure case — and it is exactly what six different domains in six weeks is for.
Many of you have your own dataset. Using it is the worst way to learn this: the context is already in your head.
Being an outsider forces the context out into the open, where the agent can read it.
Agent A behaves like a real scientist: it answers a good question honestly and a vague one uselessly. Agent B becomes whatever you configure it to be.
The transcript is where the skill is visible. We read it for:
You present the work you did in that week's lab: the problem, what you built, a critical read of the method, and what the literature says about it.
Along the way you learn the things nobody examines directly: judging whether the agent is doing the right thing, letting it run long without losing control, and turning a result into a report and a talk.
| Track | Oral presentation | Project report and labs | Course grade |
|---|---|---|---|
| Master's students | Mandatory | Pass / fail | A–F, from the oral |
| PhD students and others | Optional, welcome | Pass / fail | Pass / fail |
In six weeks you should be able to sit down with a scientist at SciLifeLab for an hour and leave with a working, checked analysis by the end of the week.
The whole toolchain is in the course repository. You are expected to work the same way, and to disclose it the same way.
What is a problem in your own lab that you would bring to a forward-deployed scientist?