PSPetr Simecek

Petr Simecek

*This is an algorithmically generated author record based on publicly available information.

Current Organizations / Institutes

The organizations/institutes from the scholar's latest article

Previous Organizations / Institutes

The organizations / institutes from the scholar's other articles in the past ten years


Petr Simecek has a total of 174 co-authors. Here are the top co-authors by number of shared publications:


Latest Publications

  • Open Access

    Advancing knotted protein design with ESM3: guided generation and topological insights

    by Eva Marsalkova, Petr Simecek

    Multimodal protein language models have transformed protein design, yet their capacity to capture complex topological features remains poorly understood. We use knotted proteins, rare structures in which the backbone forms a nontrivial topological knot, as a test case to probe this capacity using ESM3, a generative protein language model. Topology-aware guided decoding strongly enriches ESM3 outputs for knotted topologies, producing structures classified as knotted at an 89% success rate (95% CI: 81%–94%), compared to ∼0.5% for unguided diffusion-based approaches. A confidence analysis shows that freshly generated artificial knots have lower ESM3 pLDDT and pTM than real knotted proteins evaluated under the same pipeline, motivating a cautious interpretation of generated examples as model samples pending independent validation. A post hoc plausibility screen nevertheless identified a generated 31 trefoil with a near-full-length structural match to 2I6D, an experimentally determined 31-knotted SPOUT methyltransferase. In contrast, the robustness analyses on real knotted proteins are high-confidence: on average 84% of the protein sequence must be altered before the knot breaks, and the loss follows a sharp threshold rather than gradual degradation. Strikingly, structural drift accumulates well before topological disruption, suggesting that topology is more robust than specific three-dimensional arrangement. These findings position knotted proteins as a useful probe of how generative protein models represent rare, global structural features.

  • Open Access

    DNA Compression with Genomic Language Models: Tokenization, Benchmarking, and an Information-Content Map

    by Vojtech Macala, Petr Simecek

    Lossless compression and probabilistic sequence modeling are two faces of the same coin: a model that assigns high probability to a sequence can encode it in few bits via arithmetic coding. We exploit this duality to evaluate genomic language models as compressors of DNA, using compression primarily as an objective probe of generative sequence modeling rather than as a deployable storage system. We release DNAGPT2, a family of ten GPT-2-small models pretrained for one epoch on a single A40 using the DNABERT2 multi-species corpus that differ only in byte-pair encoding vocabulary size. Coupled with arithmetic coding, the best model reaches 1.47 bits per base (bpb) on the T2T human genome, fourth in the Cobilab compression benchmark and ahead of every general-purpose compressor. Our results suggest that NLP-style tokenization choices may be suboptimal for DNA: a 32-token BPE vocabulary compresses better than larger vocabularies. We also find that, in this benchmark, published long-context genomic LMs underperform a much shorter-context BPE GPT-2; we discuss in Section 5 that this is not a controlled context-length ablation, since the compared models also differ in architecture, training data, parameter count, and tokenization. Finally, we compute a per-nucleotide information-content map of the human genome and show that exons, introns, intergenic regions, and Alu repeats have statistically distinct information profiles.

  • Open Access

    Advancing Knotted Protein Design with ESM3: Guided Generation and Topological Insights

    by Eva Marsalkova, Petr Simecek

    Multimodal protein language models have transformed protein design, yet their capacity to capture complex topological features remains poorly understood. We use knotted proteins, rare structures in which the backbone forms a nontrivial topological knot, as a test case to probe this capacity using ESM3, a generative protein language model. ESM3’s guided generation produces knotted proteins with an 89% success rate (95% CI: 81–94%), compared to ∼ 0.5% for unguided diffusion-based approaches. Knot topology is remarkably robust to sequence perturbation: on average 84% of the protein sequence must be altered before the knot breaks, and the loss follows a sharp threshold rather than gradual degradation. Strikingly, structural drift accumulates well before topological disruption, suggesting that topology is more robust than specific three-dimensional arrangement. Generated proteins show no close sequence similarity to known knotted proteins, arguing against simple memorization. These findings have implications for protein engineering and, more speculatively, for discussions of biosecurity in the era of generative biological AI.

  • Top Cited Papers
    Open Access

    Balancing speed and precision in protein folding: a comparison of AlphaFold2, ESMFold, and OmegaFold

    by Anna Hýsková, Eva Maršálková, Petr Šimeček

    The rapid development of protein structure prediction tools has created a need for systematic performance comparisons to guide method selection for different applications, particularly given the trade-offs between computational speed and prediction accuracy. We benchmarked AlphaFold2, ESMFold, and OmegaFold using 1,337 protein chains deposited in the Protein Data Bank between July 2022 and July 2024, ensuring no overlap with training data, and evaluated predictions using Root Mean Square Deviation (RMSD), Template Modeling score (TM-score), Global Distance Test–Total Score (GDT-TS) and predicted Local Distance Difference Test (pLDDT) metrics. AlphaFold2 achieved the highest median TM-score (0.96), highest median GDT-TS (94%), and lowest median RMSD (1.30 Å), outperforming ESMFold (TM-score 0.95, GDT-TS 90%, RMSD 1.74 Å) and OmegaFold (TM-score 0.93, GDT-TS 89%, RMSD 1.98 Å), with all tools showing reduced accuracy for proteins lacking family annotations, leucine-rich repeats, and NMR-determined structures, while alignment-free methods unexpectedly excelled at de novo designed proteins. The performance differences between methods were negligible for many proteins, suggesting that faster alignment-free predictors (10–30 times faster) can be sufficient for numerous applications; we developed LightGBM classifiers using ProtBert embeddings and confidence scores that accurately predict when AlphaFold2’s computational investment is warranted, providing practitioners with actionable guidance for selecting between speed and precision in structural pipelines.


  • Open Access

    Balancing Speed and Precision in Protein Folding: A Comparison of AlphaFold2, ESMFold, and OmegaFold

    by Anna Hýskova, Eva Maršálková, Petr Šimeček

    We compared the performance of three widely used protein structure prediction tools—AlphaFold2, ESMFold, and OmegaFold—using a dataset of over 1,300 newly created records from the PDB database. These structures, resolved between July 2022 and July 2024, ensure unbiased evaluation, as they were unavailable during the training of these tools. Using metrics such as root mean square deviation (RMSD), template modeling score (TM-score), and predicted local distance difference test (pLDDT), we found that AlphaFold2 consistently achieves the highest accuracy but depends on high-quality sequence alignments. In contrast, ESMFold and OmegaFold provide faster predictions and excel in challenging cases, such as rapidly evolving or designed proteins with limited sequence homology. Comparing ESMFold and OmegaFold, ESMFold achieves higher confidence scores (pLDDT) and structural similarity (TM-score). OmegaFold is competitive in specific contexts, such as de novo-designed proteins or sequences with limited evolutionary information. Additionally, we demonstrate that machine learning models trained on protein language model embeddings and pLDDT confidence scores can predict potential structure prediction failures, helping to identify challenging cases early in the pipeline.


  • Open Access

    Knot or not? Identifying unknotted proteins in knotted families with sequence‐based Machine Learning model

    by Maciej Sikora, Eva Klimentova, Dawid Uchal, Denisa Sramkova, Agata P. Perlinska, Mai Lan Nguyen, Marta Korpacz, Roksana Malinowska, Szymon Nowakowski, Pawel Rubach, et al.

    Knotted proteins, although scarce, are crucial structural components of certain protein families, and their roles continue to be a topic of intense research. Capitalizing on the vast collection of protein structure predictions offered by AlphaFold (AF), this study computationally examines the entire UniProt database to create a robust dataset of knotted and unknotted proteins. Utilizing this dataset, we develop a machine learning (ML) model capable of accurately predicting the presence of knots in protein structures solely from their amino acid sequences. We tested the model's capabilities on 100 proteins whose structures had not yet been predicted by AF and found agreement with our local prediction in 92% cases. From the point of view of structural biology, we found that all potentially knotted proteins predicted by AF can be classified only into 17 families. This allows us to discover the presence of unknotted proteins in families with a highly conserved knot. We found only three new protein families: UCH, DUF4253, and DUF2254, that contain both knotted and unknotted proteins, and demonstrate that deletions within the knot core could potentially account for the observed unknotted (trivial) topology. Finally, we have shown that in the majority of knotted families (11 out of 15), the knotted topology is strictly conserved in functional proteins with very low sequence similarity. We have conclusively demonstrated that proteins AF predicts as unknotted are structurally accurate in their unknotted configurations. However, these proteins often represent nonfunctional fragments, lacking significant portions of the knot core (amino acid sequence).


  • Open Access

    Knot or Not? Sequence-Based Identification of Knotted Proteins With Machine Learning

    by Denisa Šrámková, Maciej Sikora, Dawid Uchal, Eva Klimentová, Agata P. Perlinska, Mai Lan Nguyen, Marta Korpacz, Roksana Malinowska, Pawel Rubach, Petr Šimeček, et al.

    Knotted proteins, although scarce, are crucial structural components of certain protein families, and their roles remain a topic of intense research. Capitalizing on the vast collection of protein structure predictions offered by AlphaFold, this study computationally examines the entire UniProt database to create a robust dataset of knotted and unknotted proteins. Utilizing this dataset, we develop a machine learning model capable of accurately predicting the presence of knots in protein structures solely from their amino acid sequences, with our best-performing model demonstrating a 98.5% overall accuracy. Unveiling the sequence factors that contribute to knot formation, we discover that proteins predicted to be unknotted from known knotted families are typically non-functional fragments missing a significant portion of the knot core. The study further explores the significance of the substrate binding site in knot formation, particularly within the SPOUT protein family. Our findings spotlight the potential of machine learning in enhancing our understanding of protein topology and propose further investigation into the role of knotted structures across other protein families. TOC Graphic:

  • Open Access

    Genomic benchmarks: a collection of datasets for genomic sequence classification

    by Katarína Grešová, Vlastimil Martinek, David Čechák, Petr Šimeček, Panagiotis Alexiou

    Background: Recently, deep neural networks have been successfully applied in many biological fields. In 2020, a deep learning model AlphaFold won the protein folding competition with predicted structures within the error tolerance of experimental methods. However, this solution to the most prominent bioinformatic challenge of the past 50 years has been possible only thanks to a carefully curated benchmark of experimentally predicted protein structures. In Genomics, we have similar challenges (annotation of genomes and identification of functional elements) but currently, we lack benchmarks similar to protein folding competition. Results: Here we present a collection of curated and easily accessible sequence classification datasets in the field of genomics. The proposed collection is based on a combination of novel datasets constructed from the mining of publicly available databases and existing datasets obtained from published articles. The collection currently contains nine datasets that focus on regulatory elements (promoters, enhancers, open chromatin region) from three model organisms: human, mouse, and roundworm. A simple convolution neural network is also included in a repository and can be used as a baseline model. Benchmarks and the baseline model are distributed as the Python package ‘genomic-benchmarks’, and the code is available at https://github.com/ML-Bioinfo-CEITEC/genomic_benchmarks. Conclusions: Deep learning techniques revolutionized many biological fields but mainly thanks to the carefully curated benchmarks. For the field of Genomics, we propose a collection of benchmark datasets for the classification of genomic sequences with an interface for the most commonly used deep learning libraries, implementation of the simple neural network and a training framework that can be used as a starting point for future research. The main aim of this effort is to create a repository for shared datasets that will make machine learning for genomics more comparable and reproducible while reducing the overhead of researchers who want to enter the field, leading to healthy competition and new discoveries.


  • Open Access

    Genomic Benchmarks: A Collection of Datasets for Genomic Sequence Classification

    by Katarina Gresova, Vlastimil Martinek, David Cechak, Petr Simecek, Panagiotis Alexiou

    In this paper, we propose a collection of curated and easily accessible sequence classification datasets in the field of genomics. The proposed collection is based on a combination of novel datasets constructed from the mining of publicly available databases and existing datasets obtained from published articles. The main aim of this effort is to create a repository for shared datasets that will make machine learning for genomics more comparable and reproducible while reducing the over-head of researchers that want to enter the field. The collection currently contains eight datasets that focus on regulatory elements (promoters, enhancers, open chromatin region) from three model organisms: human, mouse, and roundworm. A simple convolution neural network is also included in a repository and can be used as a baseline model. Benchmarks and the baseline model are distributed as the Python package ‘genomic-benchmarks’, and the code is available at https://github.com/ML-Bioinfo-CEITEC/genomic_benchmarks .


  • Open Access

    Author Correction: Defining the consequences of genetic variation on a proteome-wide scale

    by Joel M. Chick, Steven C. Munger, Petr Simecek, Edward L. Huttlin, Kwangbom Choi, Daniel M. Gatti, Narayanan Raghupathy, Karen L. Svenson, Gary A. Churchill, Steven P. Gygi