17.12.2025. 15:15h, Faculty of Mathematics (online)

Evolutionary costs and benefits of organismal complexity: insights from a cross-species analysis

Dr. Yury Barbitoff

Institute of Bioinformatics Research and Education and JetBrains Research, Belgrade, Serbia

A meeting of the Bioinformatics seminar will be held on Wednesday, December 17th, starting at 15:15, in online classroom. Teachers and students of doctoral and master studies in computer science, mathematics, biology and other related disciplines are invited to join us.

LINK: https://zoom.us/j/2183428158?pwd=ouAZtpLrbPnOBsKjQiarS9Rh59fyqF.1

Please note that the seminar time has been moved to Wednesday, 15:15 CET this semester.

Abstract

Reconstruction of the complex network of genotype-to-phenotype relationships is a pivotal task of modern genetics. High-throughput methods have facilitated the investigation of the network of genotype-to-phenotype relationships of the whole-phenome scale (i.e., simultaneously studying hundreds or even thousands of traits). Pleiotropy, a phenomenon of multiple phenotypic effects of the same genetic alteration, is one of the most important features of genotype-to-phenotype networks. Pleiotropy is commonly considered one of the hallmarks of organismal complexity, and theoretical models predict that it should negatively affect adaptation of an organism to the environment. We have recently conducted a series of studies aimed at dissecting the functional determinants and evolutionary consequences of pleiotropy for different species and types of traits. We found that the same mechanisms appear to be driving pleiotropic effects in different species, despite the low concordance of individual gene-trait relationships between those species. Across different types of traits, highly pleiotropic genes emerge as ones with the broadest expression profile, greatest number of protein-protein interactions, and involvement in a large number of processes. Perhaps more importantly, our results show that the high degree of pleiotropy tends to enhance both positive and negative natural selection across different species and trait domains. These results suggest that organismal complexity has both its costs and benefits for the species, and are in line with the expectations drawn from the more advanced theoretical models.

Lecturer

Yury Barbitoff is a research group leader at the Institute of Bioinformatics Research and Education (IBRE) and a Senior Researcher at JetBrains Research working in Belgrade, Serbia. At IBRE, Yury’s research group focuses on studying human genome variation, with an emphasis on systemic analysis of genotype-to-phenotype relationships and evolutionary processes shaping the landscape of genetic variation across human genes. Yury obtained his Ph.D. in Genetics from St. Petersburg State University in 2023. From 2018 to 2022, Yury has been directing research efforts at the Bioinformatics Institute (St. Petersburg, Russia), and has been the leading contributor to large-scale human population genomics endeavors, such as the construction of the RUSeq database of protein-coding genetic variation in Russia.

Seminar

The organizer of the seminar is BIRBI. The heads of the seminar are Prof. dr Nataša Pržulj and dr Jovana Kovačević.

3.12.2025. 15:15h, Faculty of Mathematics (online)

Charting the Unknown: AI-Powered Classification and Discovery Across the Protein Universe

Prof. Dr. Joana Pereira

VIB Center for AI and Computational Biology and Department Cellular and Molecular Medicine, KU Leuven, Leuven, Belgium

A meeting of the Bioinformatics seminar will be held on Wednesday, December 3rd, starting at 15:15, in online classroom. Teachers and students of doctoral and master studies in computer science, mathematics, biology and other related disciplines are invited to join us.

Please note that the seminar time has been moved to Wednesday, 15:15 CET this semester.

Abstract

The protein universe remains only partly charted, with countless families and functions yet to be uncovered. By harnessing large-scale datasets of protein sequences and structures, we recently introduced an unsupervised representation of this space that identifies functionally meaningful clusters across millions of proteins. This data-driven framework enables systematic prioritization and characterization of uncharacterized families, generating candidates for experimental validation. Through this approach, we revealed new biology at scale, including unrecognized prokaryotic defense systems and a previously undescribed repetitive protein fold, the β-flower.

Although the biological role of proteins adopting the β-flower fold remains unknown, we demonstrate that these sequences are remarkably diverse at both sequence and structural levels. They are predominantly encoded in metagenomic datasets, with expression patterns that are not yet understood. Importantly, experimentally determined structures confirm the existence of the β- flower fold, offering opportunities for engineering repetitive, soluble β-barrel–like architectures. Together, these results highlight the transformative potential of AI-guided discovery and establish the foundation for a dynamic, ever-expanding atlas of the protein universe, accelerating efforts to decode molecular functions, trace evolutionary trajectories, and inspire protein design across the tree of life.

Lecturer

Joana Pereira leads the “Protein Evolution and Function Models” group at the VIB Center for AI and Computational Biology in Leuven and is an Assistant Professor at KU Leuven’s Department of Cellular and Molecular Medicine (since February 2025). She is an expert in computational structural biology and protein evolution, with a doctorate in Chemistry from the University of Hamburg and EMBL, and a background in biochemistry. She has over seven years of postdoctoral experience at the Max Planck Institute in Tübingen and the Biozentrum in Basel. Her research integrates structural bioinformatics, complex network analysis, and deep learning to study the structure, diversity, and evolution of the protein universe, in close collaboration with experimental labs, especially on prokaryotic systems

Seminar

The organizer of the seminar is BIRBI. The heads of the seminar are Prof. dr Nataša Pržulj and dr Jovana Kovačević.

26.11.2025. 15:15h, Faculty of Mathematics (online)

Computational genomics at scale: Open-source frameworks for cancer omics and gene regulation

Prof. Dr. Aziz Khan

Mohamed bin Zayed University of Artificial Intelligence (MBZUAI), Abu Dhabi, United Arab Emirates

A meeting of the Bioinformatics seminar will be held on Wednesday, November 26th, starting at 15:15, in online classroom. Teachers and students of doctoral and master studies in computer science, mathematics, biology and other related disciplines are invited to join us.

LINK: https://zoom.us/j/2183428158?pwd=ouAZtpLrbPnOBsKjQiarS9Rh59fyqF.1

Please note that the seminar time has been moved to Wednesday, 15:15 CET this semester.

Abstract

Advances in cancer research rely on scalable and reproducible computational frameworks for analyzing genomic data. This seminar will introduce open-source tools and infrastructures for interrogating diverse cancer genomes. First, I will highlight open resources for studying gene regulation and the regulatory genome. Next, I will present the scalable bioinformatics infrastructure we built at Stanford to enable reproducible, large-scale cancer genomics, supporting initiatives such as the Human Tumor Atlas Network (HTAN) and Metastasis Research Network (MetNet). Finally, I will share our recent HTAN pre-cancer colon atlas using familial adenomatous polyposis (FAP) as a model.

Lecturer

Aziz Khan is an Assistant Professor of Computational Biology at the Mohamed bin Zayed University of Artificial Intelligence (MBZUAI) in Abu Dhabi, where he leads the Computational Biology and Cancer Regulatory Genomics (https://khanlab.bio/) Lab. His research focuses on deciphering gene regulation and the non-coding genome in cancer and precision medicine, using scalable computational and machine learning approaches. His lab develops open-source tools and integrative frameworks to interpret large-scale, multi-omics data.

Aziz earned his PhD in Bioinformatics from Tsinghua University and completed his postdoctoral research at the University of Oslo’s NCMM. Before joining MBZUAI, he was a senior research scientist at Stanford Cancer Institute, where he led core efforts in cancer genomics infrastructure and contributed to large multi-institutional initiatives, including the Human Tumor Atlas Network (HTAN) and the Metastasis Research Network (MetNet). He has developed widely used computational resources such as JASPAR, Intervene, and UniBind, and has taught and advocated for open, transparent, and reproducible science as a Carpentries Instructor and former eLife/ASAPbio Ambassador. His work bridges AI, genomics, and systems biology, with a mission to enable reproducible, collaborative, and globally impactful science.

Seminar

The organizer of the seminar is BIRBI. The heads of the seminar are Prof. dr Nataša Pržulj and dr Jovana Kovačević.