BFBela Paul Frohn

Bela Paul Frohn

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

Current Organizations / Institutes

The organizations/institutes from the scholar's latest article


Bela Paul Frohn has a total of 33 co-authors. Here are the top co-authors by number of shared publications:


Latest Publications

  • Open Access

    Reframing enzyme function prediction as conditional generation

    by William Jf Rieger, Sebastian Häussermann, Luca Herrmann, Zecheng Li, Béla P. Frohn, Manuel Maluenda, Gabriela Lobinska, Sahil Loomba, Julia C. Reisenbauer, Mikael Bodén, et al.

    Enzymes frequently exhibit promiscuous activity beyond their native roles, providing starting-points for new functions. Finding these promiscuous enzymes, especially for non-native chemical transformations, is challenging but highly valuable, as they promise novel, sustainable solutions for chemistry and biotechnology. However, current machine learning methods are poorly suited to discovering unseen chemistry as they often frame function prediction as closed set classification or a retrieval task. Here, we present Fluxion , a generative deep learning framework that learns enzymatic catalysis by modeling dynamic electron flow trajectories across the enzyme’s catalytic residues. By combining both synthetic chemistry and biochemical datasets with protein language model representations, Fluxion generates multi-step electron-flow trajectories analogous to the arrow-pushing representations used to describe enzyme reaction mechanisms. Generation is conditioned on enzyme context, including the enzyme sequence, catalytic residues, substrates, and cofactors. We show that this conditioning allows Fluxion to learn enzyme-dependent regioselectivity across cytochrome P450 enzymes with different sequences shifting the predicted reaction sites for the same substrate. We then demonstrate that Fluxion’s embeddings are useful for downstream tasks, such as specificity prediction on two experimental datasets, with and without finetuning. Finally, we show that Fluxion has the potential to transfer synthetic chemical logic to biology; it can generate the observed non-native product from real-world non-native directed evolution screens. Our results establish a proof of concept that generative modeling through mechanistic representations of enzymes can shift enzyme function prediction beyond static database retrieval and closed set classification to function generation. This conceptual framework provides a stepping stone towards an in silico generative method to discover non-native biocatalysts.

  • Open Access

    Automated synthetic cell-based screening for designed proteins with emergent functions

    by Kareem Al Nahas, Béla P. Frohn, Aleksandra Šakanović, Frank Siedler, Petra Schwille

    Designing minimal biological systems with emergent functions such as spatiotemporal self-organization is a central goal of bottom-up synthetic biology. While computational optimization and design show promise in accelerating functional protein engineering through Design-Build-Test-Learn cycles, screening libraries for complex functions remains a major challenge. Conventional screens typically lack the spatiotemporal resolution and cell-like confinement required in bottom-up synthetic biology. Here, we present PUREdrop, an automated microfluidic platform that encapsulates and expresses protein libraries in thousands of picoliter-sized synthetic cells per construct. PUREdrop distributes these across predefined wells of a 96-well plate for time-lapse imaging, enabling parallel quantification of expression kinetics and emergent functions. To demonstrate the platform’s potential, we first screen computationally re-designed variants of the bacterial cell division protein FtsZ, and identify variants with altered bundling phenotypes and distinct kinetics. We then extend our screening procedure to general protein modulators of FtsZ and identify a combination that anchors filaments to the interface, producing a ring-like phenotype. PUREdrop bridges computational protein engineering and synthetic cell research, elevating the rational engineering of complex biological function to the next level.

  • Open Access

    Automated Synthetic Cell-based Screening for Designed Proteins with Emergent Functions

    by Kareem Al Nahas, Bela Paul Frohn, Aleksandra Šakanović, Frank Siedler, Petra Schwille

    Designing minimal biological systems with emergent functions such as spatiotemporal self-organization is a central goal of bottom-up synthetic biology. While computational optimization and design show promises to accelerate functional protein engineering through Design-Build-Test-Learn cycles, screening libraries for complex functions remains a major challenge. Conventional screens typically lack the spatiotemporal resolution and cell-like confinement required in bottom-up synthetic biology. Here, we present PUREdrop, an automated microfluidic platform that encapsulates and expresses protein libraries in thousands of picolitre-sized synthetic cells per construct. These droplets are sorted into a 96-well plate and analyzed by time-lapse imaging, allowing parallel quantification of expression kinetics and emergent functions. To demonstrate the platform’s potential, we first screened computationally re-designed variants of the bacterial cell division protein FtsZ, identifying variants with improved bundling phenotypes and faster kinetics. We then extended our screening procedure towards general protein modulators of FtsZ and identified a combination that anchors filaments to the interface, producing a ring-like phenotype. PUREdrop bridges computational protein engineering and synthetic cell research, elevating the rational engineering of complex biological function to the next level.

  • ESCRTs – a multi-purpose membrane remodeling device encoded in all life forms

    by Dikla Nachmias, Béla P. Frohn, Carsten Sachse, Itzhak Mizrahi, Natalie Elia


  • Open Access

    Machine learning-aided design and screening of an emergent protein function in synthetic cells

    by Shunshi Kohyama, Béla P. Frohn, Leon Babl, Petra Schwille

    Recently, utilization of Machine Learning (ML) has led to astonishing progress in computational protein design, bringing into reach the targeted engineering of proteins for industrial and biomedical applications. However, the design of proteins for emergent functions of core relevance to cells, such as the ability to spatiotemporally self-organize and thereby structure the cellular space, is still extremely challenging. While on the generative side conditional generative models and multi-state design are on the rise, for emergent functions there is a lack of tailored screening methods as typically needed in a protein design project, both computational and experimental. Here we describe a proof-of-principle of how such screening, in silico and in vitro, can be achieved for ML-generated variants of a protein that forms intracellular spatiotemporal patterns. For computational screening we use a structure-based divide-and-conquer approach to find the most promising candidates, while for the subsequent in vitro screening we use synthetic cell-mimics as established by Bottom-Up Synthetic Biology. We then show that the best screened candidate can indeed completely substitute the wildtype gene in Escherichia coli. These results raise great hopes for the next level of synthetic biology, where ML-designed synthetic proteins will be used to engineer cellular functions.


  • Open Access

    Designing a protein with emergent function by combined in silico, in vitro and in vivo screening

    by Shunshi Kohyama, Béla P. Frohn, Leon Babl, Petra Schwille

    Recently, utilization of machine learning (ML) based methods has led to astonishing progress in protein design and, thus, the design of new biological functionality. However, emergent functions that require higher-order molecular interactions, such as the ability to self-organize, are still extremely challenging to implement. Here, we describe a comprehensive in silico, in vitro , and in vitro screening pipeline (i 3 -screening) to develop and validate ML-designed artificial homologs of a bacterial protein that confers its role in cell division through the emergent function of spatiotemporal pattern formation. Moreover, we present complete substitution of a wildtype gene by an ML-designed artificial homolog in Escherichia coli . These results raise great hopes for the next level of synthetic biology, where ML-designed synthetic proteins will be used to engineer cellular functions.


  • Open Access

    Asgard ESCRT-III and VPS4 reveal conserved chromatin binding properties of the ESCRT machinery

    by Dikla Nachmias, Nataly Melnikov, Alvah Zorea, Maya Sharon, Reut Yemini, Yasmin De-Picchoto, Ioannis Tsirkas, Amir Aharoni, Bela Frohn, Petra Schwille, et al.

    The archaeal Asgard superphylum currently stands as the most promising prokaryotic candidate, from which eukaryotic cells emerged. This unique superphylum encodes for eukaryotic signature proteins (ESP) that could shed light on the origin of eukaryotes, but the properties and function of these proteins is largely unresolved. Here, we set to understand the function of an Asgard archaeal protein family, namely the ESCRT machinery, that is conserved across all domains of life and executes basic cellular eukaryotic functions, including membrane constriction during cell division. We find that ESCRT proteins encoded in Loki archaea, express in mammalian and yeast cells, and that the Loki ESCRT-III protein, CHMP4-7, resides in the eukaryotic nucleus in both organisms. Moreover, Loki ESCRT-III proteins associated with chromatin, recruited their AAA-ATPase VPS4 counterpart to organize in discrete foci in the mammalian nucleus, and directly bind DNA. The human ESCRT-III protein, CHMP1B, exhibited similar nuclear properties and recruited both human and Asgard VPS4s to nuclear foci, indicating interspecies interactions. Mutation analysis revealed a role for the N terminal region of ESCRT-III in mediating these phenotypes in both human and Asgard ESCRTs. These findings suggest that ESCRT proteins hold chromatin binding properties that were highly preserved through the billion years of evolution separating Asgard archaea and humans. The conserved chromatin binding properties of the ESCRT membrane remodeling machinery, reported here, may have important implications for the origin of eukaryogenesis.


  • Open Access

    Tracing back variations in archaeal ESCRT-based cell division to protein domain architectures

    by Béla P. Frohn, Tobias Härtel, Jürgen Cox, Petra Schwille

    The Endosomal Sorting Complex Required for Transport (ESCRT) system is a multi-protein machinery that is involved in cell division of both Eukaryotes and Archaea. This spread across domains of life suggests that a precursor ESCRT machinery existed already at an evolutionary early stage of life, making it a promising candidate for the (re)construction of a minimal cell division machinery. There are, however, only few experimental data about ESCRT machineries in Archaea, due to high technical challenges in cultivation and microscopy. Here, we analyse the proteins of ESCRT machineries in archaea bioinformatically on a protein domain level, to enable mechanistical comparison without such challenging experiments. First, we infer that there are at least three different cell division mechanisms utilizing ESCRT proteins in archaea, probably similar in their constriction mechanisms but different in membrane tethering. Second, we show that ESCRT proteins in the archaeal super-phylum Asgard are highly similar to eukaryotic ESCRT proteins, strengthening the recently developed idea that all Eukaryotes descended from archaea. Third, we reconstruct a plausible evolutionary development of ESCRT machineries and suggest that a simple ESCRT-based constriction machinery existed in the last archaeal common ancestor. These findings not only give very interesting insights into the likely evolution of cell division in Archaea and Eukaryotes, but also offer new research avenues by suggesting hypothesis-driven experiments for both, cell biology and bottom-up synthetic biology.


  • Open Access

    Tracing back variations in archaeal cell division to protein domain architectures

    by Béla Frohn, Tobias Härtel, Jürgen Cox, Petra Schwille

    Summary: The Endosomal Sorting Complex Required for Transport (ESCRT) system is a multi-protein machinery that is involved in cell division of both Eukaryotes and Archaea. This spread across super-phyla suggests that a precursor ESCRT machinery existed already at an evolutionary early stage of life, making it a promising candidate for the (re)construction of a minimal cell division machinery. There are, however, only few experimental data about ESCRT machineries in Archaea, due to high technical challenges in cultivation and microscopy. Here, we analyse the proteins of archaeal ESCRT machineries bioinformatically on a domain level, to enable mechanistical comparison without such challenging experiments. First, we infer that there are at least three different cell division mechanisms utilizing ESCRT proteins in archaea, probably similar in their constriction mechanisms but different in membrane tethering. Second, we show that ESCRT proteins in Asgard archaea are highly similar to eukaryotic ESCRT proteins, strengthening the recently developed idea that all Eukaryotes descended from archaea. Third, we reconstruct a plausible evolutionary development of ESCRT machineries and suggest that a simple ESCRT-based constriction machinery existed in the last archaeal common ancestor. These findings not only give very interesting insights into the evolution of cell division in Archaea and Eukaryotes, but also reveal new research avenues by suggesting hypothesis-driven experiments for both, cell biology and bottom-up synthetic biology.