news – Systems Biology and Bioinformatics Group

December 9, 2025

Our Research Group at AI@DEI 2025 – From Research to Industry

Last week our research group took part in AI@DEI 2025 – From Research to Industry, the annual event organised by the Department of Information Engineering (DEI) of the University of Padova and the Regional Innovation Network IMPROVENET, with the support of UniSMART.
The initiative aims to strengthen the dialogue between academia and industry, showcasing concrete applications of Artificial Intelligence across a wide range of sectors, from computer vision to predictive maintenance, industrial decision-making, robotics, automation, and conversational technologies.

As part of the programme, our researcher Erica Tavazzi delivered the presentation: “Fondo Italiano per le Scienze Applicate: AI e diagnostica avanzata nella partnership DEI–AB Analitica contro l’antibiotico resistenza”.

The talk, prepared together with Dino Paladin (AB Analitica), introduced READY – Responsive Early Antibiotic resistance Detection and therapY, the project funded by the Italian Applied Sciences Fund (FISA) and launched this year. READY focuses on integrating AI, automation, and advanced diagnostics to support the early detection of antimicrobial resistance.

We warmly thank DEI, IMPROVENET, and UniSMART for organising the event, as well as all participating companies and colleagues for the productive exchange of perspectives.
A special acknowledgement goes to Sara Brugnerotto, whose photos beautifully captured the atmosphere of the day.

November 28, 2025

New article on biologically informed feature summarization in spatial transcriptomics now published!

🚀 We’re excited to share that our latest paper is out: Biologically Informed Procedure for Feature Summarization in Spatial Transcriptomics — a new framework designed to uncover the spatial organization of rare yet biologically crucial features, such as transcription factors, in imaging-based single-cell RNA sequencing data (e.g., MERFISH).

📊 In this work, we introduce a comprehensive procedure that integrates:
• TANGRAM, to align and integrate MERFISH with scRNA-seq data.
• Decoupler with COLLECTRI prior knowledge, to derive biologically grounded activity scores for transcription factors.
• Spatial statistics, to detect spatially informative features and highlight their organizational patterns across tissue sections.

🔍 By combining these elements, our workflow enables the discovery of subtle yet functionally relevant spatial signatures that would otherwise remain hidden — providing a powerful tool for interrogating tissue architecture in single-cell spatial genomics.

👥 This achievement was made possible thanks to the collaborative work of all authors and the support of the IJCNN organizers. Special thanks to the members of SysBioBiG – Systems Biology and Bioinformatics Group at UNIPD, who shared this journey with us.

🔗 The paper is available at: https://doi.org/10.1109/IJCNN64981.2025.11227921

November 27, 2025

New article on realistic tumoral sample simulation published in BMC Bioinformatics!

🚀 We’re excited to share that our latest paper is now published in BMC Bioinformatics: MOV&RSim: computational modelling of cancer-specific variants and sequencing reads characteristics for realistic tumoral sample simulation, https://doi.org/10.1186/s12859-025-06292-0

📊 We developed MOV&RSIM, a novel simulator that leverages data-driven information to set variants and reads characteristics, producing realistic tumoral samples, and providing full control on biological and technical parameters. Additionally, we leveraged well-annotated variant databases to create cancer-specific presets that inform the simulator’s parameters for 21 cancer types.

🔍 The proposed simulator and presets represent the most adaptable and comprehensive computational framework currently available for generating tumor samples, enabling comprehensive benchmarking and, ultimately, the optimization of somatic variant callers across diverse cancer types.

👥 This research is the result of a collaboration between our group and AB ANALITICA srl. Congratulations to the first author, Dr. Francesca Longhin, who developed MOV&RSim during her doctoral studies in our research group!

🔗 The tool is freely available on gitlab at: https://gitlab.com/sysbiobig/movarsim

October 7, 2025October 27, 2025

New article published on IEEE Access

We are pleased to announce the publication of our latest paper, “Validity of Feature Importance in Low-Performing Machine Learning for Tabular Biomedical Data”
in IEEE Access: 10.1109/ACCESS.2025.3618851

🔍 This study investigates how machine learning (ML) performance affects the validity of feature importance in biomedical datasets.
While high model accuracy is often considered a prerequisite for interpreting feature importance, this assumption has rarely been examined. In this work, we challenge this notion by showing that even low-performing models can provide reliable feature importance in biomedical contexts.

🖥️ We developed an experimental framework to assess the stability of feature importance, finding that performance degradation due to a limited number of samples behaves as conventionally expected, reducing validity, whereas degradation caused by a limited number of features preserves the validity of feature importance to a much greater extent.

👥 This research is the result of a collaboration between our group, Georgia Institute of Technology and Seoul National University. Congratulations to the first author, Dr. Youngro Lee, who was a visiting Ph.D. student in our group during his doctoral studies at Seoul National University. Great job, Youngro!

September 29, 2025

Celebrating achievements: PhD Thesis Award to Dr. Giulia Cesaro

Congratulations to our post-doc researcher Giulia Cesaro who has been awarded the PhD Thesis Award 🏆 by the Istituto di BioRobotica – Scuola Superiore Sant’Anna during the XLIV Annual School 2025 of the Gruppo Nazionale di Bioingegneria (GNB).

✨ Her doctoral dissertation was recognized by the committee “for the originality of the content and the ability to apply engineering approaches to the analysis of cell-to-cell communication from transcriptomic data, in a complex biological system, tackled with methods that reveal and model its complexity.” 💻

September 15, 2025September 15, 2025

CONVECS @ HPCSIM2025

Last week, Giacomo attended HPCSIM25 – Frontiers of High-Performance & Cloud Computing in Modeling and Simulation, held in Padova, Italy, on September 11–12, 2025.

The event focused on the latest methodologies and technologies in the field of High-Performance Computing (HPC), offering researchers and practitioners an overview of the current landscape and future directions of HPC in both real-world applications and scientific research.

Giacomo delivered an oral presentation titled “High-Performance Scientific Computing in Veneto: The CONVECS Initiative”, introducing the CONVECS project (COmuNità VEneta per il Calcolo Scientifico) where he is responsible for leading multiple core tasks. This initiative, recognized as an Operation of Strategic Importance by the Veneto Regional Government, aims to enhance, consolidate, and interconnect the HPC infrastructures currently available across universities and research laboratories in the Veneto region. Its goal is to create a unified, scalable computing environment accessible to both academic researchers and technology-driven enterprises, driving measurable impact in science, industry, and society.

September 15, 2025September 24, 2025

Sysbiobig @ CIBB2025

Last week we attended the 20th Conference on Computational Intelligence Methods for Bioinformatics and Biostatistics (CIBB2025) in Milano from September 10th to 12th, 2025.

We’re proud to share that our group contributed to the scientific program with three oral presentations showcasing our recent research efforts:

🧠 Piero Mariotto, Ilaria Patuzzi, Giada Innocente, Barbara Simionati, Barbara Di Camillo, Giacomo Baruzzo “A network science-based approach to unveil the effects of faecal microbiota transplantation in enteropathic dogs” (great collaboration with EuBiome srl)

📊 Federico De Mori Bajolin, Anna Maria Bianchi, Erica Tavazzi “A multivariate deep-learning approach for stratifying Amyotrophic Lateral Sclerosis patients based on temporal dynamics“

💉 Davide Dei Cas, Barbara Di Camillo, Gian Paolo Fadini, Giovanni Sparacino, Enrico Longato “The impact of clinical history on the predictive performance of machine learning and deep learning models for renal complications of diabetes“

July 29, 2025July 29, 2025

Sysbiobig @ISMB/ECCB 2025

Last week, Barbara, Giacomo, Giulia, Gaia, and Matteo attended the 33rd Conference on Intelligent Systems for Molecular Biology & 24th European Conference on Computational Biology (ISMB/ECCB 2025), held in Liverpool, UK, from July 20–24, 2025.

The conference was a valuable opportunity to attend cutting-edge presentations, engage in discussions on the latest advances in bioinformatics, and foster new collaborations.

We’re proud to share that our group contributed to the scientific program with several presentations:

Giulia Cesaro delivered an oral presentation in the NetBio session titled “Cell-specific Graph Operation Strategy on Signaling Intracellular Pathways” showcasing work done in collaboration with the CostaLab at RWTH Aachen University.
Giacomo Baruzzo presented the poster “Realistic Simulation of NGS Reads from Tumoral Samples with MOV&RSim”, a project led by Francesca Longhin (former PhD student of our group), in collaboration with AB Analitica.
Matteo Baldan presented the poster “Integrating Biological Knowledge for Feature Summarization in Spatial Transcriptomics”.
Gaia Tussardi presented the poster “Multilevel Network Visualization for Deciphering Dysregulated Cellular Signalling”.
In collaboration with the CostaLab at RWTH Aachen University, we also co-organized the tutorial “Computational Approaches for Deciphering Cell-Cell Communication from Single-Cell and Spatial Transcriptomics Data”.

July 9, 2025

Our paper is out in NAR Genomics and Bioinformatics

🚀 We’re excited to share that our latest paper is now published in NAR Genomics and Bioinformatics: Differential cellular communication inference framework for large-scale single-cell RNA-sequencing data, https://doi.org/10.1093/nargab/lqaf084

We introduce a novel computational framework tailored for analyzing and interpreting differential cell–cell communication in complex, large-scale scRNA-seq datasets. The framework incorporates two tools: scSeqCommDiff, which identifies and characterizes alterations in cell–cell communication in a fast and memory-efficient manner, and CClens, which facilitates interpretation and exploration through an interactive R/Shiny interface.ì

What’s new?
👥 Works across diverse experimental designs
🔍 Captures both intercellular and intracellular signaling
📊 Designed for big data: fast and memory-efficient
🖥️ Comes with an interactive Shiny app for easy and insightful exploration

Explore the tools:
🔗 https://gitlab.com/sysbiobig/scseqcomm
🔗 https://gitlab.com/sysbiobig/cclens

July 7, 2025

Our contribution to AIME 2025 in Pavia

Last week, our group attended the 23^rd International Conference on Artificial Intelligence in Medicine (AIME 2025) in Pavia, Italy.

It was a fantastic opportunity to follow high-quality presentations and engage with leading researchers and professionals from across the AI and healthcare communities.

We are proud to share that we contributed to the scientific program with four research papers:

Davide Dei Cas held a short oral presentation entitled “Exploring the Use of Projecting Conflicting Gradients in Multi-Task Neural Networks with an Application to Amyotrophic Lateral Sclerosis.”

Erica Tavazzi co-authored the poster “Towards Distributed Process Discovery in Healthcare: Testing and Proving the Feasibility of the Federated Alpha+ Algorithm.”

At the 2^nd International Workshop on Process Mining Applications for Healthcare (PM4H 2025), our group was also involved in two contributions: “Predicting Next Clinical Event in Amyotrophic Lateral Sclerosis using Process-Oriented Machine Learning Models: a Case Study” and “Federated I-PALIA: Privacy-By-Design Distributed Process Discovery for Duplicated Activities in Healthcare” (🏆Best Paper Award – PM4H 2025).

We sincerely thank the conference organizers for hosting this event and for giving us the opportunity to share our research.