Automation of research processes with LLMs and AI

Venue

Eva & Georg Klein lecture hall, Biomedicum
Solnavägen 9, 171 65 Solna
Solna, 171 65 Sweden

Automation of research processes with LLMs and AI

March 6, 2025 @ 10:00 15:00 CET

SciLifeLab is thrilled to invite you to the symposium ‘Automation of research processes with LLMs and AI’, to be held on March 6 at Eva & Georg Klein lecture hall, Biomedicum, Karolinska Institutet Campus Solna.

The purpose of the symposium is practical knowledge sharing, inspiration, and reflection on the future of scientific processes in life sciences and health research. Academic, healthcare, and industry researchers working within the broad life sciences are welcome to attend.

More and more AI-based tools are currently being integrated in scientific research processes, from simple academic text editing tools to automated labs empowered by large language models (LLMs) running experiments and making scientific conclusions. For researchers in life sciences and health, developments in the AI technologies are important because of the high potential, large monetary investments, but also expectations about future breakthroughs.

This one-day event offers an opportunity for the researchers to hear about specific examples of incorporating Large Language Models (LLMs) and other recent developments in foundational AI models in research workflows as well as to discuss the future of scientific research processes with this new technology present.

We will start the morning with inspirational talks about new advances in research workflow automation. During lunch, there will be a possibility to network and discuss different perspectives with other attendees. After lunch we plan to have interactive sessions where the audience will have an opportunity to share and discuss visions and expectations, including spontaneous flash/pop-up talks by the attendees.

The talks of the morning session will be livestreamed through Zoom. The rest of the event is on-site in Biomedicum only.

SPEAKERS AND PROGRAMME TO BE ANNOUNCED SHORTLY

Registration

Registration

Participation is free of charge on the first-come first-served basis. Refreshments and lunch will be served for the on-site participants. Unfortunately, we cannot accommodate allergies or dietary preferences for those who register after March 1. If you register to listen to livestreamed talks on Zoom, you will receive a link on the day of the event. The talks will not be recorded and saved, only streamed live on the day.

Waiting List

To avoid empty seats, registration will remain open until the event begins. If we reach full capacity, a waiting list will be activated. Sign up for the waiting list, and you will automatically receive an email when a spot becomes available. You must accept to secure the spot. If you decline, the offer will go to the next person on the waiting list.

Cancellation

To minimize empty seats and especially food waste, you must cancel your registration if you are unable to attend the event. If a waiting list is activated, your spot will go to someone else. Otherwise, your cancellation will result in us covering the cost of your refreshments and lunch.

If you have questions, please contact datacentre@scilifelab.se

Eva & Georg Klein lecture hall

Speakers

Nobel prizes, huge tech investments, generative AI and AI agents, multidisciplinary expertise, expectations on emerging superhuman intelligence, and 1,000+ AI-powered approved medical devices all highlight AI’s transformative potential for life scientists. Yet, many researchers remain unaware or skeptical, often due to limited exposure. This talk will cover AI’s current and future impact in life sciences, highlighting how it can empower every scientist, boost research productivity, and foster innovative approaches.

Olli Kallioniemi is the Founding Director of the SciLifeLab & Wallenberg Data-Driven Life Science (DDLS) Program and served as Director of SciLifeLab from 2015 to 2024. He is a Professor of Molecular Precision Medicine at Karolinska Institute (OnkPat) and holds a joint appointment at the Institute for Molecular Medicine Finland (FIMM) at the University of Helsinki. He has recently focused on exploring and explaining the opportunities that AI offers for life science research and how AI can benefit every researcher.

Europe PubMed Central (Europe PMC), developed by EMBL-EBI, is a free, open-access literature database containing over 45 million life science literature outputs; over 10 million of which are full text. This includes almost a million preprints from 34 different servers. Associated research outputs, such as data, are linked to the literature, which is also enriched with additional persistent identifiers, text mined funding information, and text mined annotations of the articles and supplementary files. Annotations are dictionary-based and are being transitioned to a machine learning approach. This talk will discuss the potential of this transition and will explore how Europe PMC plans to leverage machine learning in the future while maintaining the community’s trust in its curated, open dataset as a reliable resource for scientific research. Words of caution about AI and ghost workers, including biocurators, will also be mentioned.

Melissa Harrison is the Lead of Literature Services at EMBL’s European Bioinformatics Institute (EMBL-EBI). After completing a biology degree at King’s College London she worked in scientific publishing for two decades. In her previous role at eLife she honed her skills in open science policy and implementation and has been instrumental in the propagation of open persistent identifiers and metadata.

Aasa Feragen is a Professor in Medical Image Analysis at the Technical University of Denmark (DTU).

Foundation models are undoubtedly highly useful. But are they a panacea for advancing performance of deep learning systems? If not, what are their limitations? In the presentation, we explore these questions using large and rich digitized prostate pathology data as an example.

Martin Eklund is professor of epidemiology at the Department of Medical Epidemiology and Biostatistics where he focusses his research on reducing the mortality of prostate cancer as well as the negative side effects of today’s imprecise diagnostics and treatment selection. He works a lot on development of AI systems for clinical decision making.

Data-driven science has emerged as a modern pillar of science, complementing theory, experiments and computational approaches. Scientific workflows – from design of experiments to data analysis, have greatly benefited from tools and techniques from various fields including optimization, statistics, computational science, etc. In this talk, we will motivate principled approaches for various stages of the scientific research pipeline, from traditional foundations such as optimization, to recent AI-based advances such as large-language models.

Prashant is an Assistant Professor, and a SciLifeLab fellow hosted at the Department of Information Technology, Science for Life Laboratory, Uppsala University. His research interests involve developing machine learning and optimization methods to enable fast, data-efficient analysis and processing of scientific data, particularly in the domain of life sciences.

Nextflow has become the leading tool for scientific workflows in bioinformatics, but writing workflows can be challenging. This talk introduces Seqera AI, a new tool that helps generate Nextflow pipelines either from scratch, from existing scripts, or even other workflow languages. I’ll discuss how Seqera AI is able to generate best-practice DSL2 code that follows nf-core guidelines, and how we go beyond just code generation: creating and running test suites with AI agents, streamlining development while improving code quality and validation.

Phil Ewels is Senior Product Manager for Open Source at Seqera, working with Nextflow, MultiQC, nf-core and Wave. Before joining Seqera in 2022, Phil worked at the National Genomics Infrastructure (NGI) at SciLifeLab in Stockholm, Sweden. It was through this work that Phil became involved with Nextflow and eventually co-founded the nf-core community. Phil’s career has spanned many disciplines from lab work and bioinformatics research in epigenetics, through to software development and community engagement. He is the author of MultiQC, SRA-Explorer, QCFail.com, and has a PhD in Molecular Biology from the University of Cambridge, UK.

Program

TBA

Solnavägen 9, 171 65 Solna
Solna, 171 65 Sweden

Last updated: 2025-02-10

Content Responsible: Arnold Kochari(arnold.kochari@scilifelab.uu.se)