Machine Learning-Guided Protein Engineering

ACS Catalysis, Год журнала: 2023, Номер 13(21), С. 13863 - 13895

Опубликована: Окт. 13, 2023

Recent progress in engineering highly promising biocatalysts has increasingly involved machine learning methods. These methods leverage existing experimental and simulation data to aid the discovery annotation of enzymes, as well suggesting beneficial mutations for improving known targets. The field protein is gathering steam, driven by recent success stories notable other areas. It already encompasses ambitious tasks such understanding predicting structure function, catalytic efficiency, enantioselectivity, dynamics, stability, solubility, aggregation, more. Nonetheless, still evolving, with many challenges overcome questions address. In this Perspective, we provide an overview ongoing trends domain, highlight case studies, examine current limitations learning-based We emphasize crucial importance thorough validation emerging models before their use rational design. present our opinions on fundamental problems outline potential directions future research.

Язык: Английский

---

Overlay databank unlocks data-driven analyses of biomolecules for all

Nature Communications, Год журнала: 2024, Номер 15(1)

Опубликована: Фев. 7, 2024

Abstract Tools based on artificial intelligence (AI) are currently revolutionising many fields, yet their applications often limited by the lack of suitable training data in programmatically accessible format. Here we propose an effective solution to make scattered various locations and formats for data-driven machine learning using overlay databank To demonstrate practical relevance such approach, present NMRlipids Databank—a community-driven, open-for-all database featuring programmatic access quality-evaluated atom-resolution molecular dynamics simulations cellular membranes. Cellular membrane lipid composition is implicated diseases controls major biological functions, but membranes difficult study experimentally due intrinsic disorder complex phase behaviour. While MD have been useful understanding systems, they require significant computational resources suffer from inaccuracies model parameters. Here, how programmable interface flexible implementation applications, rapid simulation through a graphical user interface, unlock possibilities beyond current experimental studies understand The proposed concept can be further applied other biomolecules, as well fields where similar barriers hinder AI revolution.

Язык: Английский

---

One bead per residue can describe all-atom protein structures

Structure, Год журнала: 2023, Номер 32(1), С. 97 - 111.e6

Опубликована: Ноя. 23, 2023

Язык: Английский

---

From complex data to clear insights: visualizing molecular dynamics trajectories

Frontiers in Bioinformatics, Год журнала: 2024, Номер 4

Опубликована: Апрель 11, 2024

Advances in simulations, combined with technological developments high-performance computing, have made it possible to produce a physically accurate dynamic representation of complex biological systems involving millions billions atoms over increasingly long simulation times. The analysis these computed simulations is crucial, the interpretation structural and data gain insights into underlying processes. However, this becomes challenging due complexity generated large number individual runs, ranging from hundreds thousands trajectories. This massive increase raw creates additional processing visualization challenges. Effective techniques play vital role facilitating molecular dynamics simulations. In paper, we focus mainly on tools that can be used for among which highlight few approaches specifically purpose, discussing their advantages limitations, addressing future challenges visualization.

Язык: Английский

---

Permeability Benchmarking: Guidelines for Comparing in Silico, in Vitro, and in Vivo Measurements

Journal of Chemical Information and Modeling, Год журнала: 2025, Номер unknown

Опубликована: Янв. 17, 2025

Permeability is a measure of the degree to which cells can transport molecules across biological barriers. Units permeability are distance per unit time (typically cm/s), where accurate measurements needed define drug delivery in homeostasis and model dysfunction occurring during disease. This perspective offers set community-led guidelines benchmark data multidisciplinary approaches different contexts. First, we lay out analytical framework for three methodologies calculate permeability: silico assays using either transition-based counting or inhomogeneous-solubility diffusion approaches, vitro cultured 2D 3D geometries, vivo utilizing situ brain perfusion multiple time-point regression analysis. Then, demonstrate systematic benchmarking both vivo, depicting ways each sensitive choices assay design. Finally, outline seven recommendations best practices underscore significance tailored driving advancements research development. Our exploration encompasses discussion "generic" tissue-specific barriers, including blood–brain barrier (BBB), major hurdle therapeutic agents into brain. By addressing challenges reconciling simulated with experimental assays, aim provide insights essential optimizing accuracy reliability modeling.

Язык: Английский

---

Modeling membranes in situ

Current Opinion in Structural Biology, Год журнала: 2024, Номер 87, С. 102837 - 102837

Опубликована: Май 13, 2024

Molecular dynamics simulations of cellular membranes have come a long way-from simple model lipid bilayers to multicomponent systems capturing the crowded and complex nature real cell membranes. In this opinionated minireview, we discuss current challenge simulate in their native environment, situ, with prospect reaching level whole cells organelles using an integrative modeling framework.

Язык: Английский

---

Probing the dynamic landscape of peptides in molecular assemblies by synergized NMR experiments and MD simulations

Communications Chemistry, Год журнала: 2024, Номер 7(1)

Опубликована: Фев. 13, 2024

Peptides or proteins containing small biomolecular aggregates, such as micelles, bicelles, droplets and nanodiscs, are pivotal in many fields ranging from structural biology to pharmaceutics. Monitoring dynamics of systems has been limited by the lack experimental methods that could directly detect their fast (picosecond nanosecond) timescale dynamics. Spin relaxation times NMR experiments sensitive motions, but interpretation for aggregates is not straightforward. Here we show dynamic landscape peptide-containing molecular assemblies can be determined a synergistic combination solution state (MD) simulations. Solution straightforward implement without an excessive amount sample, while direct spin data MD simulations enables landscapes peptides other aggregated molecules. To demonstrate this, interpret transmembrane, peripheral, tail anchored embedded micelles. Our results indicate detergent molecules do rotate together rigid body, viscous medium composed micelle. also provide indirect information on peptide conformational ensembles. This work gives new perspectives complex assemblies.

Язык: Английский

---

mdciao: Accessible Analysis and Visualization of Molecular Dynamics Simulation Data

PLoS Computational Biology, Год журнала: 2025, Номер 21(4), С. e1012837 - e1012837

Опубликована: Апрель 21, 2025

We present mdciao, an open-source command line tool and Python Application-Programming-Interface (API) for easy, one-shot analysis representation of molecular dynamics (MD) simulation data. Building upon the widely used concept residue-residue contact-frequencies, mdciao offers a wide spectrum further representations, enriched with available domain specific annotations. The user-friendly interface pre-packaged solutions non-expert users, while keeping customizability expert ones. Emphasis has been put into automatically producing annotated, production-ready figures tables. Furthermore, seamless on-the-fly query inclusion domain-specific generic residue numbering GPCRs, GAIN-domains, G-proteins, kinases is made possible through online lookups. This allows easy selection comparison across different systems, regardless sequence identity, target residues or domains. Finally, fully documented API users to include basic advanced functions in their workflows, provides numerous examples Jupyter Notebook Tutorials. source code published under GNU Lesser General Public License v3.0 later hosted on https://github.com/gph82/mdciao , documentation, including guides examples, can be found at https://www.mdciao.org

Язык: Английский

---

OGT Enhances Adriamycin Resistance of Breast Cancer by Promoting Glycolysis through MDM4 Upregulation in an O-GlcNAcylation-Dependent Manner

Biochemical Genetics, Год журнала: 2025, Номер unknown

Опубликована: Май 15, 2025

Язык: Английский

---

Advancing Molecular Dynamics: Toward Standardization, Integration, and Data Accessibility in Structural Biology

The Journal of Physical Chemistry B, Год журнала: 2024, Номер 128(10), С. 2219 - 2227

Опубликована: Фев. 28, 2024

Molecular dynamics (MD) simulations have become a valuable tool in structural biology, offering insights into complex biological systems that are difficult to obtain through experimental techniques alone. The lack of available data sets and structures most published computational work has limited other researchers' use these models. In recent years, the emergence online sharing platforms MD database initiatives favor deposition ensembles accompany publications, favoring reuse sets. However, uniform metadata collection, formats, what deposited limits impact its by different communities not necessarily experts MD. This Perspective highlights need for standardization better resource processing interpreting simulation results, akin efforts areas biology. As field moves forward, we will see an increase popularity benefits MD-based integrative approaches combining probabilistic reasoning, but too uniformity availability choices on how modeled trivial decipher from papers. Other fields addressed similar challenges comprehensively establishing task forces with degrees success. large scope number represent breadth types complicates parallel approach would fit all. Thus, each group typically decides which format upload servers like Zenodo. Uploading FAIR (findable, accessible, interoperable, reusable) principles mind including optimal collection make more accessible actionable community. Such wealth foster method development infrastructure advancements, thus propelling forward.

Язык: Английский

---