BBraun Medical Kft. Budapest, HU | SOFTWARE DEVELOPER iNTERN | 2022‐06 ‐ 2022‐08

• Contributed to the development of software for acute dialysis machines.

• Conducted unit testing to ensure software quality and reliability

• Programming experience in C/C++

CISCO Systems Hungary Budapest, HU | DEVOPS ENGiNEER iNTERN | 2021‐04 ‐ 2022‐01

• Engaged in networking and API programming activities

• Supporting DevOps practices and infrastructure

Background:
Analyzing large-scale, mass spectrometry–based complex proteomics datasets often overwhelms desktop computational resources and requires substantial manual configuration. While FragPipe provides rapid peptide identification across diverse sample preparation and acquisition modes (DDA, DIA, TMT), deploying it at scale remains challenging.

Results:
We introduce Frag’n’Flow, a Nextflow-based pipeline that encapsulates FragPipe, automates input manifest creation and workflow generation, manages tool dependencies, and includes downstream data analysis options. This enables reproducible, high-performance analysis on HPC, cloud, and cluster environments. Benchmarking against other workflow-based solutions demonstrates that Frag’n’Flow maintains quantitative accuracy while reducing runtime by nearly half on a typical DIA dataset (~58 GB), alleviating memory and I/O bottlenecks. We further validate Frag’n’Flow across three representative datasets—label-free DDA, DIA, and TMT—successfully recapitulating published biological signatures with minimal user intervention.

Conclusions:
By combining the sensitivity and speed of FragPipe with Nextflow’s orchestration capabilities, Frag’n’Flow enables scalable analysis of large proteomics datasets and empowers researchers, regardless of computational expertise, to extract deeper insights from existing MS datasets.

Availability: Frag’n’Flow is available at: https://github.com/ronalabrcns/FragNFlow

TurboID-based proximity labeling is a powerful approach to capture protein-protein interactions within their native cellular environment. Here, we present a step-by-step protocol for fusing proliferating cell nuclear antigen (PCNA) to TurboID and generating stable cell lines via lentiviral transduction. We describe steps for cell synchronization, DNA damage induction, and proximity labeling, followed by fractionation, affinity purification, and mass spectrometry to identify biotinylated proteins. 

For complete details on the use and execution of this protocol, please refer to Rona et al. 2024. 

Background: In the evolving landscape of microbiology and microbiome analysis, the integration of machine learning is crucial for understanding complex microbial interactions, and predicting and recognizing novel functionalities within extensive datasets. However, the effectiveness of these methods in microbiology faces challenges due to the complex and heterogeneous nature of microbial data, further complicated by low signal-to-noise ratios, context-dependency, and a significant shortage of appropriately labeled datasets. This study introduces the ProkBERT model family, a collection of large language models, designed for genomic tasks. It provides a generalizable sequence representation for nucleotide sequences, learned from unlabeled genome data. This approach helps overcome the above-mentioned limitations in the field, thereby improving our understanding of microbial ecosystems and their impact on health and disease.

Methods: ProkBERT models are based on transfer learning and self-supervised methodologies, enabling them to use the abundant yet complex microbial data effectively. The introduction of the novel Local Context-Aware (LCA) tokenization technique marks a significant advancement, allowing ProkBERT to overcome the contextual limitations of traditional transformer models. This methodology not only retains rich local context but also demonstrates remarkable adaptability across various bioinformatics tasks.

Results: In practical applications such as promoter prediction and phage identification, the ProkBERT models show superior performance. For promoter prediction tasks, the top-performing model achieved a Matthews Correlation Coefficient (MCC) of 0.74 for E. coli and 0.62 in mixed-species contexts. In phage identification, ProkBERT models consistently outperformed established tools like VirSorter2 and DeepVirFinder, achieving an MCC of 0.85. These results underscore the models' exceptional accuracy and generalizability in both supervised and unsupervised tasks.

Conclusions: The ProkBERT model family is a compact yet powerful tool in the field of microbiology and bioinformatics. Its capacity for rapid, accurate analyses and its adaptability across a spectrum of tasks marks a significant advancement in machine learning applications in microbiology. The models are available on GitHub (https://github.com/nbrg-ppcu/prokbert) and HuggingFace (https://huggingface.co/nerualbioinfo) providing an accessible tool for the community.

2025

• Oral presentation  - Hungarian Society for Bioinformatics Conference, Budapest HU

• Poster presentation  - Central and Eastern European Proteomics Conference, Budapest HU (poster competition award)

• Poster presentation - Central European Genome Stability Meeting, Debrecen HU

• Poster presentation - Semmelweis University Scientific Days Conference, Budapest HU (best poster award Molecular Medicine)

• Poster presenation -  Hungarian Biotechnology Student Association Biotech Days, Budapest HU

2025

• Proteomics Bioinformatics Course 2025 - EMBL EBI, Hinxton UK

2025

• University Researchers’ Scholarship Program Cooperative Doctoral Program (EKÖP-KDP) - 36 months

• Pannonia Scholarships - Travel Grant

• European Proteomics Association - Travel Grant 2026

2024

• University Researchers’ Scholarship Program (EKÖP) - 12 months

• Hungarian Bioinformatics Society (2025- )

• Hungarian Neuroscience Society (2025- )

• Hungarian Biochemichal Society (2024- )

• Scientific Association for Infocommunications (2022- )

• Mensa HungarIQa (2017- )