🧬 KTH Royal Institute of Technology · SciLifeLab
Data-Driven
Life Sciences
A hands-on course where you master the AI tools, data science workflows, and biological reasoning skills that are reshaping how science is done — and how researchers stay competitive.
🤖 AI Agents & LLMs
🔬 Microscopy & Imaging
🧬 Single-cell Genomics
🧪 Protein Structure
💊 Precision Medicine
🔭 Systems Biology
7.5
ECTS Credits
6
Modules
100+
Students Trained
4×
Years Running
⚡ Why This Matters Now
You're Graduating Into a Different World
Every decade, a shift in tools changes how knowledge workers operate. The current shift is the biggest yet — and it's happening right now.
📚 How we find information
📚
Pre-2000
Library & Textbooks
🔍
2000s
Google Search
💬
2022+
ChatGPT & LLMs
🤖
Today
AI Agents & Autonomous Research
💻 How we write code
⌨️
Pre-2020
Type every line in an IDE
✨
2021
GitHub Copilot suggests lines
⚡
2023
Cursor writes entire functions
🚀
Today
AI agents code, test & debug entire projects
The right question isn't "will AI replace scientists?"
It's: will scientists who use AI effectively replace those who don't?
This course gives you that edge — how to prompt correctly, direct AI agents for data analysis,
train models, build tools, and critically evaluate AI outputs in a scientific context.
🔬 The Changing Nature of Science
How Biological Research Has Evolved
The scientific method itself is being transformed — from a purely human-driven process to one where AI plays an increasingly central role in every step.
Traditional Science
Pre-2000s → ongoing
Hypothesis-Driven Research
A scientist forms a hypothesis based on prior knowledge and intuition, designs a targeted experiment to test it, and interprets results manually. Powerful — but slow, narrow, and limited by human capacity to process data.
Observe phenomenon
Form hypothesis
Design & run experiment
Collect & analyze data manually
Publish & repeat
e.g.
Koch's postulates, Mendel's genetics, classical drug trials — one hypothesis at a time.
Computational Biology
2000s → present
Data-Driven Hypothesis Generation
Large-scale biological datasets (genomes, proteomes, single-cell profiles) are analyzed computationally. Machine learning reveals patterns that generate new hypotheses — often ones a human would never have considered.
Generate massive dataset (omics, imaging)
Apply ML / statistical models
Discover unexpected patterns
Form data-driven hypothesis
Validate experimentally
e.g.
GWAS reveals disease-linked variants across 500,000 genomes; scRNA-seq uncovers unknown cell types; AlphaFold predicts protein structure from sequence alone.
🚀 Emerging Now
2024 → future
The AI Scientist — Autonomous Research
AI agents autonomously review literature, generate hypotheses, design experiments, write and execute code, interpret results, and iterate — completing in hours what took researchers months. Humans set the goals; AI drives the loop.
AI reviews entire literature corpus
Generates & ranks hypotheses
Designs & simulates experiments
Analyzes results, updates model
Drafts manuscript autonomously
e.g.
Sakana AI's "AI Scientist" writes & reviews its own papers; AI discovers novel antibiotics (MIT, 2023); automated lab robots close the experiment loop end-to-end.
🔬 What is Data-Driven Life Sciences?
Where Biology Meets Computation
Modern biology generates more data than any human can process. Data-driven life sciences is the discipline of using machine learning, AI, and computational methods to extract meaning from this data — and to make discoveries that would be impossible otherwise.
AlphaFold
DeepMind's AI solved protein structure prediction — a 60-year grand challenge — in 2020. Every protein in the human genome, predicted in days.
Structural BiologySingle-Cell Sequencing
We can now profile the gene expression of individual cells, mapping every cell type in the human body at unprecedented resolution.
GenomicsAI Microscopy
Deep learning can now detect cancer cells, classify organisms, and measure cellular dynamics from microscopy images with superhuman accuracy.
ImagingPrecision Medicine
By integrating genomic, clinical, and environmental data, AI is enabling treatments tailored to individual patients rather than disease averages.
Systems BiologySix Modules, Six Fields
Each week brings a new domain, taught by leading researchers from SciLifeLab and KTH.
1
Intro to Data-Driven Life Sciences
Foundations of AI in biology, using AI agents and LLMs for scientific exploration.
2
Image Analysis & Microscopy
Deep learning for biological images: cell segmentation, classification, super-resolution.
3
Protein Structure & Molecular Biology
AlphaFold, protein language models, molecular simulation, and structure prediction.
4
Single-cell Transcriptomics & Genomics
scRNA-seq analysis, dimensionality reduction, trajectory inference, spatial transcriptomics.
5
Precision Medicine & Systems Biology
Multi-omics integration, network biology, patient stratification, drug target discovery.
6
Automated Scientific Discovery & AI Agents
AI agents that design experiments, analyze results, and drive autonomous scientific workflows.
Weekly Lecture by Leading Researcher
Hands-on AI-Augmented Lab (Google Colab)
Journal Club & Peer Discussion
Final Project: Build Something Real
🎯 How You'll Learn
Learning, Augmented by AI
Every week combines expert knowledge with AI-assisted hands-on practice — the same workflow you'll use in your research career.
Expert Lectures
Weekly talks by DDLS Fellows, SciLifeLab facility leaders, and KTH researchers using cutting-edge methods.
2 hrs/weekAI-Augmented Labs
Hands-on Google Colab notebooks with AI coding assistants. You direct the analysis; AI handles the boilerplate.
4 hrs/weekJournal Club
Critically read, present, and debate landmark papers in data-driven life sciences. AI helps you prepare.
2 hrs/weekReal Final Project
Build a working tool, web app, or analysis pipeline. Past students built apps that got deployed in real labs.
3 weeks🎬 Student Projects
What Past Students Built
The final project is 3 weeks of building something real — AI-powered tools, web apps, and research demos used by actual labs.
🔬🧬
🌊 Environmental Science · SciLifeLab
Karin Garefelt
Built a full-stack web app for classifying plankton from IFCB flow cytometry images — deployed at SciLifeLab and now used by researchers.
Watch Presentation🤖💻
⚡ AI Coding Workflows · KTH PhD
Lasse Stahnke
Developed an AI-assisted coding workflow tool that uses agents to automate data analysis pipelines — from raw data to publishable figures.
Watch Presentation🧪🔭
🔬 Research Automation · SciLifeLab
Augusta Jensen
Created an AI-driven research demo showing how an AI agent can search literature, summarize findings, and propose next experimental steps.
Watch PresentationExplore top courses
.js-id-current
📊 HT 2025: Data-Driven Life Sciences
Course content for the 2025 Data-Driven Life Sciences (DDLS) course.
What Students Say
“
The DDLS course broadened my general knowledge in data-driven life sciences. Throughout the course, we were learning how to use AI tools effectively. I was surprised by how quickly I could build my first web app!
“
The course provided a solid introduction to applying agentic AI in both research and coding. The final project was especially useful as a chance to put the AI-supported coding workflows into practice.
“
This course was a great introduction to the world of data-driven life science research. It contained interesting lectures as well as hands-on experience working with Google Colab and Gemini CLI as an AI agent.
Meet the Team
Wei Ouyang
Teacher / Examiner
Nils Mechtel
Teaching Assistant
