AI in bioinformatics – where Life Science meets Computer Science

We are happy to announce our 8th UBC symposium.

The UBC symposium will be held on Monday, October 10, 2022 in Gasthuis Leeuwenbergh, Servaasbolwerk 1a, Utrecht.​ Please join us for this unique symposium that is bringing together bioinformaticians and computer scientists who are accelerating AI development in the field of bioinformatics.

We will also celebrate the opening of our GPU facility by highlighting the technology and use cases. This has been made possible by the AI4UU initiative, a special investment of 610k€ from Utrecht University to expand the GPU capabilities of our HPC infrastructure, in order to strengthen its compute facility in AI and deep learning

We strongly encourage all members of the UBC and beyond to register for the symposium (free registration) and to apply for a presentation. Please send an abstract (max 300 words) before Thursday, September 29 to Corine Heuzer ( specifying if you would like to present orally and/or with a poster. We have digital boards (no printing needed, only a USB stick)!

Please note that the then-applicable corona rules will be followed.

UBC Symposium October 10, 2022: "AI in bioinformatics" image

Monday, October 10 2022

Gasthuis Leeuwenbergh | Servaasbolwerk 1a | 3512 NK | Utrecht


Send your abstract (300 words max, deadline Wednesday, September 28)


Final program



08.30 – 09.00 Registration
09.00 – 09.15 Welcome by Dr. Jeroen de Ridder (UMCU, executive board UBC)
09:15 – 10:00 Keynote speaker: Dr. Ahmed Mahfouz (LUMC) – “Learning cellular response to perturbations from single-cell and spatial genomic data”
10:00 – 10:30 Coffee break
10:30 – 11:00 Lucía Barbadilla Martínez (UMCU) – “Prediction of variant effects on non-coding regions with deep learning models”
11:00 – 11:30 Dr. Timothy Dallman (UU) – “Prediction of clinical outcome of Escherichia coli O157:H7 infection using machine learning”
11:30 – 12:00 Dr. Ies Nijman (UMCU) – “Presentation of the AI4UU initiative” & Dr. Nikolas Stathonikos (UMCU) – “AI/ML in clinical practice – Histopathology perspective”
12:00 – 13:30 Walking lunch and Poster Presentations
13:30 – 14:15 Keynote speaker: Dr. Jonas Teuwen (NKI) – “Predicting weak labels in high-dimensional histopathology data”
14:15 – 14:45 Petros Skiadas (UU) – “Studying the highly variable effector repertoire of a major crop pathogen using structural genomics”
14:45 – 15:15 Marc Pagès-Gallego (UMCU) – “Detection of 8-oxodG using Nanopore sequencing and neural networks”
15:15 – 15:45 Dr. Sjors Middelkamp (PMC) – “Tracing the origin of pediatric cancer using single-cell whole genome sequencing at unprecedented resolution”
15:45 – 16:15 Coffee break
16:15 – 17:00 Keynote speaker: Prof. dr. Lodewyk Wessels (NKI) – “Understanding cellular decision making”
17:00 – 18:00 Closing ceremony followed by drinks
initial image
initial image

Sara Pulit | Bringing genetics to drug discovery: What computational biology can (and can’t do – yet) to help find new medicines

from Vertex Pharmaceuticals, San Diego: Sara has been working in Utrecht, with a strong emphasis on genetic association studies to unravel genetic predispositions for different diseases. She has recently moved to Vertex Pharmaceuticals where she continues to apply genetics and bioinformatics in an industrial setting.

Abstract: The mapping of the human genome nearly two decades ago has resulted in a tidal wave of genetic discoveries. Linkage analyses revealed the underpinnings of rare diseases, genome-wide association studies yielded insights into common diseases, and sequencing continues to uncover the genetic keys to an array of syndromes and disorders. With this veritable mountain of discoveries before us, a primary aim of genetics (and computational biology more broadly) is to discover if and how these findings can lead us to transformative medicines for serious diseases. In my talk, I will describe how genetic studies – performed even before the human genome had been mapped – paved the way to our understanding of cystic fibrosis (CF) and in turn informed existing therapeutics for the treatment of CF. I will also discuss how genetics and computational biology has transformed and continues to transform the drug discovery process.

Lorem ipsum

Vlad Cojocaru | Structural modeling and molecular simulations of protein-DNA interactions

I will present here our recent efforts in understanding the mechanisms by which transcription factors recognize DNA in different chromatin contexts. Transcription factors are proteins that regulate gene expression by binding to DNA regulatory regions to activate or repress the transcription of genomic DNA into RNA. Many transcription factors are involved in determining the identity of a cell and usually just a few of them are sufficient to convert between different cellular states. Such conversions have important applications in regenerative therapies but are often inefficient and uncontrollable. Understanding the mechanisms by which transcription factors recognize their sequence specific binding sites on DNA and especially on DNA wrapped in chromatin will provide means to optimize cell state conversions. I will demonstrate how molecular simulations can be used to decode these mechanisms.

Anita Schürch | Prediction and exploration of the pan-plasmidome of medically relevant bacterial species 

In bacteria, a considerable part of the genome can be encoded on mobile genetic elements. These mobile genetic elements disseminate important traits such as antibiotic resistance genes which can spread through different hosts and environments. However, the reconstruction of mobile genetic elements such as plasmids from short read genomic and metagenomic sequencing data is challenging. I will present novel computational strategies to reconstruct plasmids and their potential to study plasmid sequences at the bacterial species level. This allowed the  exploration of the pan-plasmidome of Enterococcus faecium, an important multi-drug resistant nosocomial pathogen. We observed that isolates from hospital patients carry more plasmid sequences than isolates from other sources. Moreover, plasmidomes rather than chromosomes are highly specific for the isolation source.