About this Research Topic
In silico modeling tools allow us to guide protein engineering and improve our fundamental understanding of biocatalysis. Bioinformatics, statistical modeling, machine learning and deep learning provide the tools to broaden the space of possible biochemical reactions by facilitating the generation of more efficient or de novo biocatalysts and optimizing the modeling of enzymatic reaction cascades and metabolic pathways. Statistical and Artificial Intelligence (AI) approaches, fed by experimental data, helps us to decipher how the sequence (and internal interactions), structure and dynamics of proteins, taken alone or in combination, determine the protein function and the associated fitness landscape. Protein engineering in the context of metabolic engineering has a growing impact in industrial biotechnology and synthetic biology.
However, deciphering protein sequence-structure-function relationships remains a major challenge in protein chemistry and enzymology, particularly when non-linear phenomena take place. Ignorance of the subtleties of enzymatic mechanisms and the impact of single or multiple mutations in the protein sequence and final function can make this challenge even more difficult when using rational design or directed evolution predictive approaches. Generating robust models of cascades for enzymatic reactions further comprises an unmet goal. Epistasis occurs when a combination of mutations exhibits significantly different fitness compared to the sum of their independent effects.
Understanding the evolution of proteins cannot be done without the knowledge of the epistasis phenomena (non-additive mutational effects) that take place within them. Epistasis can indeed reverse the effect of a mutation from beneficial to deleterious. Likewise, the conservation of a neutral mutation during evolution can lead, thanks to epistasis phenomena, to beneficial effects of greater amplitude. Interactions between residues of mutated amino acids, as well as intramolecular interaction networks that can be set up following mutations, condition the protein function. While it is clear that understanding sequence-structure-function relationships, and in particular intramolecular epistasis, is of paramount importance in protein engineering, the means to predict these phenomena are currently limited. Non-linear interactions are still poorly understood, making the design of networks of interacting amino acid residues to introduce desired functionality in an enzyme a bottleneck.
The aim of this Research Topic is to give an overview of recent advances and discuss the current understanding of epistatic phenomena in protein engineering. Particular attention will be paid to the better understanding and modeling of epistatic phenomena that can impair the prediction of a property of interest associated with an enzyme. Similarly, contributions on predictive models, which are a challenge in the field, will be particularly appreciated. These include models that:
i. are interpolative, extrapolative and can predict out-of-the-box (i.e. outside the range of values learned during the modeling phase),
ii. can operate simultaneously on large or small training data sets,
and
iii. can fully capture non-linear interactions in the protein sequence.
Contributions in the fields described below are particularly, but not exclusively, welcome:
• State of the art computer simulations and protein engineering techniques to enhance our fundamental understanding of enzymes, their mechanisms and evolution at the atomistic level
• Transdisciplinary scientific methods for developing novel enzyme engineering strategies for a broad range of applications: biopolymer, drugs, fine chemicals
• Characterization of multienzyme complexes (stoichiometry, kinetic properties and regulation),
• Identification of protein-protein interactions
• AI (Machine Learning & Deep Learning) applied to Protein engineering and Modeling of metabolic pathways in cellular or cell-free systems
Prof. Cadet is linked to Peaccel and declares competing interests. The other Topic Editors declare no competing interests with regards to the Research Topic.
However, deciphering protein sequence-structure-function relationships remains a major challenge in protein chemistry and enzymology, particularly when non-linear phenomena take place. Ignorance of the subtleties of enzymatic mechanisms and the impact of single or multiple mutations in the protein sequence and final function can make this challenge even more difficult when using rational design or directed evolution predictive approaches. Generating robust models of cascades for enzymatic reactions further comprises an unmet goal. Epistasis occurs when a combination of mutations exhibits significantly different fitness compared to the sum of their independent effects.
Understanding the evolution of proteins cannot be done without the knowledge of the epistasis phenomena (non-additive mutational effects) that take place within them. Epistasis can indeed reverse the effect of a mutation from beneficial to deleterious. Likewise, the conservation of a neutral mutation during evolution can lead, thanks to epistasis phenomena, to beneficial effects of greater amplitude. Interactions between residues of mutated amino acids, as well as intramolecular interaction networks that can be set up following mutations, condition the protein function. While it is clear that understanding sequence-structure-function relationships, and in particular intramolecular epistasis, is of paramount importance in protein engineering, the means to predict these phenomena are currently limited. Non-linear interactions are still poorly understood, making the design of networks of interacting amino acid residues to introduce desired functionality in an enzyme a bottleneck.
The aim of this Research Topic is to give an overview of recent advances and discuss the current understanding of epistatic phenomena in protein engineering. Particular attention will be paid to the better understanding and modeling of epistatic phenomena that can impair the prediction of a property of interest associated with an enzyme. Similarly, contributions on predictive models, which are a challenge in the field, will be particularly appreciated. These include models that:
i. are interpolative, extrapolative and can predict out-of-the-box (i.e. outside the range of values learned during the modeling phase),
ii. can operate simultaneously on large or small training data sets,
and
iii. can fully capture non-linear interactions in the protein sequence.
Contributions in the fields described below are particularly, but not exclusively, welcome:
• State of the art computer simulations and protein engineering techniques to enhance our fundamental understanding of enzymes, their mechanisms and evolution at the atomistic level
• Transdisciplinary scientific methods for developing novel enzyme engineering strategies for a broad range of applications: biopolymer, drugs, fine chemicals
• Characterization of multienzyme complexes (stoichiometry, kinetic properties and regulation),
• Identification of protein-protein interactions
• AI (Machine Learning & Deep Learning) applied to Protein engineering and Modeling of metabolic pathways in cellular or cell-free systems
Prof. Cadet is linked to Peaccel and declares competing interests. The other Topic Editors declare no competing interests with regards to the Research Topic.
Keywords: epistasis, structure and/or sequence function relationships, Artificial intelligence, Machine learning, Deep Learning, bioinformatics, rational design, directed evolution, modeling, cascade reactions
Important Note: All contributions to this Research Topic must be within the scope of the section and journal to which they are submitted, as defined in their mission statements. Frontiers reserves the right to guide an out-of-scope manuscript to a more suitable section or journal at any stage of peer review.
