Welcome speech by Béla Merkely, Rector of Semmelweis University and Roland Molontay, the Director of the Organizing Comittee.

Methods for generating spatial omics data are going through a fast technology development, and they are being applied to exploring the tissue microenvironment in health and disease. However, the analysis of this high-dimensional data is challenging. I will discuss strategies for overcoming these challenges including self-supervised learning approaches to learn the structure of the data, and methods for building simple models of the data and evaluating these models using simulations. I will show examples of how these two complementary strategies, one modeling the data using a large number of parameters and the second one using few parameters, can be applied to understanding the tissue microenvironment.

• Bertold Pál: Integrating CT Radiomics and Proteomics to Characterize Small-Cell Lung Cancer

• Shreya Ghosh: Clinical Response Prediction for CFTR Modulator Therapy by Integrating Longitudinal Airway Metagenomics and Secretomics in an Age-Stratified Cohort

• Csongor Szaniszló: Segmentation of Cardiac Structures on CT Scans

Due to the availability of data and corresponding computing capacities, more and more cognitive tasks can be assigned to computers that learn independently to improve our understanding, increase our problem-solving capacity or simply help us remember. In this way, artificial intelligence (AI) systems are finding more and more niches in which they can improve our capabilities not only physically but also intellectually. In recent years, we focused on a systematic collection of data to get a deep understanding of biological entities and the relations among them. Starting with the classification and current developments in the field of AI, this talk will give an overview of how the data is collected, curated and employed. Specifically, how such data can be used to develop practical applications for biological sequence analyses, how to combine multiple data modalities to identify disease-causing biomarkers, and how to develop explainable applications for drug repurposing and repositioning.

• József Pintér: Estimating ROC-AUC for Clinical Scores under Range Restriction

• Tamás Kói: Meta-analysis of medians and quartiles

• Ubaida Fatima: Handling Uncertainty in Biomedical Data: A Fuzzy Machine Learning Approach for Disease Prediction

• Krisztián Buza: Enhanced ROCKET for the Automated Detection of Epileptic Tonic-Clonic Seizures Using Accelerometer Data

• Onur Yolay: From Prototype to Clinical Translation: Data-Driven Validation of an Adjustable Surgical Plate System for Orthognathic Surgery

• Luca Laura Bogner: NeoTracker: A Real-time Digital Event Log and Voice-Controlled Data Acquisition System for Neonatal Resuscitation and Delivery Room Care

Session Chair: Artúr Vlaszov

• Umut Aygün: EEG-Based Decoding of Pinch Force Intention: Evaluating Machine and Deep Learning Models for Advanced Brain-Computer Interfaces 

• Elmira Bagautdinova: Reproducing bistable dynamics of two-dimensional model of a Hodgkin-Huxley-type physiological neuron using ML-methods 

• Attila Barna: Graph representation of bacterial cell-wall architecture from AFM: network extraction, interpretable damage descriptors, and patch-level readouts in E. coli 

• Zoltán Bereczki: Artificial-intelligence-based identification of biomarkers for long COVID subtypes 

• Eszter Birtalan: Soft body simulation for robotics in healthcare and prosthetics 

• Flóra Demeter: Impact of Genetic Background, Passage Number, and Confluency on the Transcriptome of Mesenchymal Stem Cells 

• Péter Domján: Uncertainty-Aware Path Selection in Graphs Using an Adapted Imprecision Entropy Indicator 

• Gobena Woldemariam Erkalo: Multilevel mixed effect analysis of factors associated with practice of sunlight exposure of infants among mothers in Mettu District, Illubabor zone, Oromia Region, Ethiopia

• Aleksandr Gonchenko: Complex Oscillatory Activity in Neuron Models Manifested as the Formation of Singular Shilnikov Attractors

• Miaoxin Huang: Integrated transcriptome analysis with mendelian randomization screening for biomarkers associated with prediabetes

• Benedetta Ingino: Echoes of Longevity: Analysis of Healthy Ageing Discourse on Social media, News and Scientific literature 

• Csaba Kiss: The Critical Role of Model Selection in Causal Inference: A Comparative Analysis of Classification Models within the InferBERT Framework for Pharmacovigilance 

• Gábor Koós: Chronic beta-blocker therapy in patients after cardiac arrest: mortality and catecholamine requirement 

• Donát Köller: Exposures Influencing the Performance of Health Industry Start-Ups and Spin-offs: a Systematic Review and Meta-Analysis 

• Pauline Maniapi: Digital Health and Personalized Medicine Advantages and Future Challenges for Healthcare Industries in Cameroon 

• Anett Nagy-Szakolczai: Beyond Diagnosis and Sex: Transdiagnostic Symptom Profiles Characterize Functional Heterogeneity in ASD, ADHD, and AuDHD 

• Bertold Pál:  Zero-Shot Deep Learning For Paleopathology: Spinal Segmentation In Archaeological CT via Totalspineseg

• Ádám Pál-Jakab: Meteorological Associations with Out-of-Hospital Cardiac Arrest: A National Time-Series Analysis in Hungary 

• Anton Panyushev: A computational method for multistability detection in neuron model analysis 

• Lea Seep: Assessment of the gut microbiome in age-related macular degeneration 

• Balázs Szabó:  Multiomic Characterization of Acute and Post-COVID Patients Using Chest X-Ray Radiomics and Proteomics 

• Szandra Szedlacsek: The beneficial effect of cognitive-motor training after stroke 

• Anett Nagy-Szakolczai: Derivation of a Simple Rule-Out Tool for Occult Lymph Node Invasion in PSMA PET/CT-Negative Prostate Cancer Patients: A Retrospective Registry Analysis

• Balázs Szigeti: Integrated Risk Stratification and Clinical Reporting Platform for Pediatric Acute Leukemias

• Tibor Kézi: NEVADA: National Evidence-based Validation using AI for Disease Analysis—An In Silico Framework for Adapting Clinical Trial Evidence to a National Real-World Population

• Ádám Pál-Jakab: Multimodal Decision Support Across the Out-of-Hospital Cardiac Arrest Care Pathway: From EMS Data to Clinical Action

• Kíra Diána Kovács: Integrative Machine Learning Analysis of Epigenetic, Microbiome, and Fitness Data for Biological Age Prediction

• Laura Bagi: Revealing the Contribution of Vaccine-Related Immune Genes to Lung Cancer Predisposition via Ultra-Rare Protein Truncating Variants

• Eperke Merkel: ARIANA: an AI-based ECG risk stratification platform for early prediction of atrial fibrillation and sudden cardiac death

Virtually all processes in health and disease rely on the careful orchestration of a large number of diverse individual components ranging from molecules to cells and entire organs. Networks provide a powerful framework for describing and understanding these complex systems in a holistic fashion. They offer a unique combination of a highly intuitive, qualitative description, and a plethora of analytical, quantitative tools. In my presentation, I will review how molecular networks can be understood as maps for elucidating the relation between molecular-level perturbations and their phenotypic manifestations. I will then sketch out a number of challenges in the areas of network biology and network medicine, as well as recent efforts of my group to address them. These efforts range from methodological work on the visualization and interpretation of large biomedical data combining artificial intelligence with virtual reality technology, to translational efforts towards concrete clinical applications in rare diseases and drug repurposing.

• Ubaida Fatima: Dangling Centrality: Identifying Critical Proteins in Biomedical Networks through Link Removal Analysis

• Csaba Kiss: Collaboration, Citation Impact, and Thematic Evolution in Pharmacology: Evidence from a 75-year Bibliometric Analysis of Q1 Journals

• Jelena Smiljanić: Inferring context-specific modules in biological systems

• Zita Zarándy: Individuality and information content of infrared molecular profiles: insights from a large longitudinal health-profiling study

• Li-Lan Chuang: Learning Analytics of LINE-Based Communication Platforms Supporting Nursing Clinical Internship Training: A Data-Driven Comparative Study

• Saeed Sani: Depressive symptoms and sleep quality in IBD patients and university students: a cross-sectional pilot study

• Deborah L. Nichols: From Latent Classes to Explainable Caregiver Personalization: A Translational Digital Health Model

Venue

The Gala Dinner of the 2nd Biomedical Data Science Conference will be held at the exclusive Urban Betyár. Located in the heart of Budapest at 1051 Budapest, Október 6. utca 16–18, this elegant venue offers a unique blend of Hungarian culinary tradition and modern ambiance. 

How does brain function emerge from the organisation of complex neural systems? What network features enable flexible, adaptive computation across scales? And when these principles are disrupted during development, how do their effects cascade from genes to circuits to behaviour?

This keynote approaches these questions from a systems-level, network-based perspective. Neurodevelopmental disorders provide a powerful test case for this approach: they are intrinsically multiscale, spanning genetic risk, altered developmental trajectories, large-scale network reorganisation, and distributed cognitive and behavioural symptoms. No single level of explanation is sufficient. The central challenge is to trace causal threads across scales and to understand how biological perturbations reshape developing brain networks and their computational capacity.

Drawing on advances in connectomics, genomics, computational modelling, and network science, this talk will explore emerging efforts to bridge these levels of description. By joining the dots from molecules to networks to behaviour, we move closer to a mechanistic account of brain organisation in health and disease - one that may ultimately support more principled biomarkers and therapeutic strategies for neurodevelopmental psychiatric disorders.

• Szandra Szedlacsek: The impact of visual information on cortical activity in healthy and epileptic mice

• Balázs Szabó: Applying state-of-the-art approximate inference method on parameter inference problem connected to neurons

• Tibor Nánási: Chronos longitudinal plasma proteomics reveals biologically coherent pre-onset molecular changes in Parkinson’s disease

• Bálint File: Disentangling Biological and Non-Biological Drivers of Proteomic Signals in the Chronos-PD Study for Robust Biomarker Discovery in Parkinson's Disease