Добірка наукової літератури з теми "AlphaFold"
Оформте джерело за APA, MLA, Chicago, Harvard та іншими стилями
Ознайомтеся зі списками актуальних статей, книг, дисертацій, тез та інших наукових джерел на тему "AlphaFold".
Біля кожної праці в переліку літератури доступна кнопка «Додати до бібліографії». Скористайтеся нею – і ми автоматично оформимо бібліографічне посилання на обрану працю в потрібному вам стилі цитування: APA, MLA, «Гарвард», «Чикаго», «Ванкувер» тощо.
Також ви можете завантажити повний текст наукової публікації у форматі «.pdf» та прочитати онлайн анотацію до роботи, якщо відповідні параметри наявні в метаданих.
Статті в журналах з теми "AlphaFold":
1
Finkelstein, Alexei V. "Protein 3D Structure Identification by AlphaFold: a Physics-Based Prediction or Recognition Using Huge Databases?" Journal of Molecular Biology 6, no. 1 (March 20, 2024): 1–10. http://dx.doi.org/10.52338/tjomb.2024.3935.
Анотація:
The great success of AlphaFold programs poses the questions: (i) What is the main reason for this success? (ii) What AlphaFolds does: physics-based prediction of the spatial structure of a protein from its amino acid sequence or recognition of this structure from similarity of the target sequence to some parts of sequences with already known spatial structures? The answers given here are: (i) the main reason for the AlphaFold’s success is the usage of huge databases which already cover virtually all protein superfamilies existing in Nature; (ii) using these databases, multiple sequence alignments, and coevolutionary information – like correlations of amino acid residues of the contacting chain regions – AlphaFold recognizes a spatial structure by similarity of the target sequence (or its parts) to related sequence(s) with already known spatial structures. We emphasize that this does not diminish the merit and utility of AlphaFold but only explains the basis of its success.
2
Wheeler, Richard John. "A resource for improved predictions of Trypanosoma and Leishmania protein three-dimensional structure." PLOS ONE 16, no. 11 (November 11, 2021): e0259871. http://dx.doi.org/10.1371/journal.pone.0259871.
Анотація:
AlphaFold2 and RoseTTAfold represent a transformative advance for predicting protein structure. They are able to make very high-quality predictions given a high-quality alignment of the protein sequence with related proteins. These predictions are now readily available via the AlphaFold database of predicted structures and AlphaFold or RoseTTAfold Colaboratory notebooks for custom predictions. However, predictions for some species tend to be lower confidence than model organisms. Problematic species include Trypanosoma cruzi and Leishmania infantum: important unicellular eukaryotic human parasites in an early-branching eukaryotic lineage. The cause appears to be due to poor sampling of this branch of life (Discoba) in the protein sequences databases used for the AlphaFold database and ColabFold. Here, by comprehensively gathering openly available protein sequence data for Discoba species, significant improvements to AlphaFold2 protein structure prediction over the AlphaFold database and ColabFold are demonstrated. This is made available as an easy-to-use tool for the parasitology community in the form of Colaboratory notebooks for generating multiple sequence alignments and AlphaFold2 predictions of protein structure for Trypanosoma, Leishmania and related species.
3
Wen, Haosheng, Wei-Hsiang Weng, and Marcos Sotomayor. "A curated AlphaFold 2/AlphaFill cadherinosome." Biophysical Journal 122, no. 3 (February 2023): 474a—475a. http://dx.doi.org/10.1016/j.bpj.2022.11.2544.
4
Manabe, Noriyoshi. "AlphaFill: Docking Ligands and Cofactors into AlphaFold Models." Trends in Glycoscience and Glycotechnology 35, no. 206 (July 25, 2023): E61. http://dx.doi.org/10.4052/tigg.2316.6e.
5
AlQuraishi, Mohammed. "AlphaFold at CASP13." Bioinformatics 35, no. 22 (May 22, 2019): 4862–65. http://dx.doi.org/10.1093/bioinformatics/btz422.
Анотація:
Abstract Summary: Computational prediction of protein structure from sequence is broadly viewed as a foundational problem of biochemistry and one of the most difficult challenges in bioinformatics. Once every two years the Critical Assessment of protein Structure Prediction (CASP) experiments are held to assess the state of the art in the field in a blind fashion, by presenting predictor groups with protein sequences whose structures have been solved but have not yet been made publicly available. The first CASP was organized in 1994, and the latest, CASP13, took place last December, when for the first time the industrial laboratory DeepMind entered the competition. DeepMind's entry, AlphaFold, placed first in the Free Modeling (FM) category, which assesses methods on their ability to predict novel protein folds (the Zhang group placed first in the Template-Based Modeling (TBM) category, which assess methods on predicting proteins whose folds are related to ones already in the Protein Data Bank.) DeepMind's success generated significant public interest. Their approach builds on two ideas developed in the academic community during the preceding decade: (i) the use of co-evolutionary analysis to map residue co-variation in protein sequence to physical contact in protein structure, and (ii) the application of deep neural networks to robustly identify patterns in protein sequence and co-evolutionary couplings and convert them into contact maps. In this Letter, we contextualize the significance of DeepMind's entry within the broader history of CASP, relate AlphaFold's methodological advances to prior work, and speculate on the future of this important problem.
6
Varadi, Mihaly, Stephen Anyango, Mandar Deshpande, Sreenath Nair, Cindy Natassia, Galabina Yordanova, David Yuan, et al. "AlphaFold Protein Structure Database: massively expanding the structural coverage of protein-sequence space with high-accuracy models." Nucleic Acids Research 50, no. D1 (November 17, 2021): D439—D444. http://dx.doi.org/10.1093/nar/gkab1061.
Анотація:
Abstract The AlphaFold Protein Structure Database (AlphaFold DB, https://alphafold.ebi.ac.uk) is an openly accessible, extensive database of high-accuracy protein-structure predictions. Powered by AlphaFold v2.0 of DeepMind, it has enabled an unprecedented expansion of the structural coverage of the known protein-sequence space. AlphaFold DB provides programmatic access to and interactive visualization of predicted atomic coordinates, per-residue and pairwise model-confidence estimates and predicted aligned errors. The initial release of AlphaFold DB contains over 360,000 predicted structures across 21 model-organism proteomes, which will soon be expanded to cover most of the (over 100 million) representative sequences from the UniRef90 data set.
7
Pak, Marina A., Karina A. Markhieva, Mariia S. Novikova, Dmitry S. Petrov, Ilya S. Vorobyev, Ekaterina S. Maksimova, Fyodor A. Kondrashov, and Dmitry N. Ivankov. "Using AlphaFold to predict the impact of single mutations on protein stability and function." PLOS ONE 18, no. 3 (March 16, 2023): e0282689. http://dx.doi.org/10.1371/journal.pone.0282689.
Анотація:
AlphaFold changed the field of structural biology by achieving three-dimensional (3D) structure prediction from protein sequence at experimental quality. The astounding success even led to claims that the protein folding problem is “solved”. However, protein folding problem is more than just structure prediction from sequence. Presently, it is unknown if the AlphaFold-triggered revolution could help to solve other problems related to protein folding. Here we assay the ability of AlphaFold to predict the impact of single mutations on protein stability (ΔΔG) and function. To study the question we extracted the pLDDT and <pLDDT> metrics from AlphaFold predictions before and after single mutation in a protein and correlated the predicted change with the experimentally known ΔΔG values. Additionally, we correlated the same AlphaFold pLDDT metrics with the impact of a single mutation on structure using a large scale dataset of single mutations in GFP with the experimentally assayed levels of fluorescence. We found a very weak or no correlation between AlphaFold output metrics and change of protein stability or fluorescence. Our results imply that AlphaFold may not be immediately applied to other problems or applications in protein folding.
8
Le Page, Michael. "Why AlphaFold is transformational." New Scientist 255, no. 3398 (August 2022): 11. http://dx.doi.org/10.1016/s0262-4079(22)01373-2.
9
Laura Howes. "AlphaFold goes all atom." C&EN Global Enterprise 101, no. 37 (November 13, 2023): 6. http://dx.doi.org/10.1021/cen-10137-scicon3.
10
Porta-Pardo, Eduard, Victoria Ruiz-Serra, Samuel Valentini, and Alfonso Valencia. "The structural coverage of the human proteome before and after AlphaFold." PLOS Computational Biology 18, no. 1 (January 24, 2022): e1009818. http://dx.doi.org/10.1371/journal.pcbi.1009818.
Анотація:
The protein structure field is experiencing a revolution. From the increased throughput of techniques to determine experimental structures, to developments such as cryo-EM that allow us to find the structures of large protein complexes or, more recently, the development of artificial intelligence tools, such as AlphaFold, that can predict with high accuracy the folding of proteins for which the availability of homology templates is limited. Here we quantify the effect of the recently released AlphaFold database of protein structural models in our knowledge on human proteins. Our results indicate that our current baseline for structural coverage of 48%, considering experimentally-derived or template-based homology models, elevates up to 76% when including AlphaFold predictions. At the same time the fraction of dark proteome is reduced from 26% to just 10% when AlphaFold models are considered. Furthermore, although the coverage of disease-associated genes and mutations was near complete before AlphaFold release (69% of Clinvar pathogenic mutations and 88% of oncogenic mutations), AlphaFold models still provide an additional coverage of 3% to 13% of these critically important sets of biomedical genes and mutations. Finally, we show how the contribution of AlphaFold models to the structural coverage of non-human organisms, including important pathogenic bacteria, is significantly larger than that of the human proteome. Overall, our results show that the sequence-structure gap of human proteins has almost disappeared, an outstanding success of direct consequences for the knowledge on the human genome and the derived medical applications.
Дисертації з теми "AlphaFold":
1
Gosset, Simon. "Annotation du proteome d'A. thaliana via l'analyse et la prédiction de son interactome." Electronic Thesis or Diss., université Paris-Saclay, 2024. http://www.theses.fr/2024UPASL004.
Анотація:
Le fonctionnement des cellules vivantes est assuré par un ensemble d'interactions entre des molécules que l'on appelle protéines. Identifier les couples de protéines en interaction, impliqués dans un processus biologique d'intérêt, permet donc de mieux comprendre son fonctionnement. Pour cela, il existe un ensemble de méthodes expérimentales, mais qui sont trop coûteuses pour explorer l'ensemble des interactions mises en jeu ou pas passez fiables. Pour pallier à ce problème, des méthodes informatiques ont été développées. L'objectif de ma thèse a ainsi été de mettre en place une méthode permettant de construire un réseau d'interactions protéine-protéine (PPI) impliquant des protéines d'intérêt, puis d'y rechercher toutes les protéines impliquées dans un même processus biologique qui constituent des sous-réseaux de protéines fortement in-terconnectées. Le résultat dépendant de la qualité du réseau, j'ai tenté ensuite de l'affiner en y ajoutant des PPI prédites. Pour cela, j'ai mis en place une méthode de prédiction de PPI utilisant Alphafold-multimer, une méthode innovante de prédiction de structure de complexe protéique. En parallèle, j'ai produit un jeu de données décrivant les caractéristiques physico-chimiques, l'énergie de liaison et la propension à l'interaction des surfaces et des interfaces d'un grand nombre de protéines afin de comprendre ce qui distingue les PPI correctement prédites de celles ratées par la méthode que j'ai mise en placeLiving organisms function thanks to a set of interactions between molecules called proteins. Identifying the pairs of proteins involved in these interactions provides a better understanding of how they function. There are several experimental methods for identifying these proteins, but they are either too expensive or not reliable enough to explore all the protein interactions that take place in living organisms. Computational methods have been developed to predict these interactions to meet this need. During my thesis, I set up a method for predicting protein-protein interactions using Alphafold-multimer, an innovative structure prediction method. At the same time, I produced a dataset describing the physico-chemical characteristics, binding energy and interaction propensity of the surfaces and interfaces of a large number of proteins to understand what distinguishes correctly predicted PPIs from incorrectly predicted PPI by the method I have set up
2
Aimeur, Sana. "Exploration de l'interaction p22phox/p47phox : une clé pour comprendre l'assemblage du complexe NADPH oxydase." Electronic Thesis or Diss., université Paris-Saclay, 2023. http://www.theses.fr/2023UPASF099.
Анотація:
Le complexe NADPH oxydase phagocytaire (NOX2), occupe une place cruciale dans la génération d'espèces réactives de l'oxygène, jouant un rôle essentiel dans l'immunité innée, ainsi que dans divers processus physiologiques, quant à sa dérégulation participe aux processus de stress oxydatif, du vieillissement ou de l'inflammation. Ce complexe enzymatique complexe est constitué de six sous-unités, dont deux sous unités membranaires, NOX2 et p22phox, tandis que les quatre autres sont des protéines cytosoliques (p47phox, p67phox, p40phox et Rac). NOX2 et p22phox, codés respectivement par les gènes CYBB et CYBA, forment le cœur catalytique du complexe NADPH oxydase des phagocytes. Ce cœur, également appelé flavocytochrome b558 (Cytb558), contient tous les intermédiaires rédox permettant les transferts d'électrons nécessaire à la réduction de l'oxygène en anions superoxyde. Malgré les défis persistants tel que la flexibilité du complexe, la compréhension de la structure et des fonctions de ces sous-unités est un domaine de recherche en croissance. La première étape de l'activation de l'enzyme nécessite l'interaction entre la protéine p47phox et la sous-unité membranaire p22phox. L’exploration de ces deux protéines, et qui plus est celle de leur interaction, sont rendues difficiles en raison des leurs propriétés intrinsèques (régions désordonnées, domaines membranaires, organisation en domaine d’une grande flexibilité). Bien que plusieurs études aient menées, Les mécanismes moléculaires mis en jeu restent à clarifier. Cette thèse se concentre sur plusieurs aspects de cette interaction en abordant cette problématique en combinant des techniques biochimies des protéines solubles et membranaires) avec des études par dichroïsme circulaire et par résonance magnétique nucléaire (RMN), des mesures de diffusion des rayons X aux petits angles (SAXS) et des prédictions de structure par intelligence artificielle (AlphaFold). Ces approches complémentaires ont permis de proposer un scénario montrant les changements conformationnels de p47phox de son état inactif vers des états actifs ont pu être démontrés. Ces changements de structure de p47phox ont par ailleurs été explorés dans le cadre de son interaction avec p22phox. Cela a révélé des aspects structuraux de cette interaction jusque-là méconnu. Parallèlement, l’intelligence artificielle au travers de l'outil AlphaFold a permis d’apporter pour la première fois des modèles structuraux du complexe enzymatique assemblé. Cela a permis d’explorer les interactions protéine-protéine et protéine-membrane lipidique qui devraient encourager des études futures plus approfondies. Enfin, la mise au point de la production et la purification de la protéine membranaire p22phox recombinante, notamment en nanodisques natifs, apporte à ce travail mixant approches expérimentales et théoriques des perspectives nouvellesThe phagocytic NADPH oxidase complex (NOX2) plays a pivotal role in generating reactive oxygen species, crucial for innate immunity and various physiological processes. Its dysregulation contributes to oxidative stress, aging, and inflammation. This complex enzyme comprises six subunits, including two membrane subunits, NOX2 and p22phox, and four cytosolic proteins (p47phox, p67phox, p40phox, and Rac). NOX2 and p22phox, encoded by the CYBB and CYBA genes, respectively, form the catalytic core of the phagocyte NADPH oxidase complex. This core, also known as flavocytochrome b558 (Cytb558), contains all redox intermediates for electron transfers needed to reduce oxygen to superoxide anions. Despite ongoing challenges like the complex's flexibility, understanding the structure and function of these subunits is an expanding research area. The enzyme's activation first requires the interaction between the p47phox protein and the p22phox membrane subunit. Investigating these proteins, particularly their interaction, is challenging due to their intrinsic properties, such as disordered regions, membrane domains, and highly flexible domain organization. Although several studies have been conducted, the molecular mechanisms involved remain to be clarified. This thesis focuses on several aspects of this interaction, addressing the issue by combining biochemical techniques (for both soluble and membrane proteins) with circular dichroism studies, nuclear magnetic resonance (NMR), small-angle X-ray scattering (SAXS), and structure predictions by artificial intelligence (AlphaFold). These complementary approaches have enabled the proposal of a scenario demonstrating the conformational changes of p47phox from its inactive state to active states. These structural changes of p47phox have also been explored in the context of its interaction with p22phox, revealing previously unknown structural aspects of this interaction. Concurrently, artificial intelligence through AlphaFold has provided the first structural models of the assembled enzyme complex, allowing exploration of protein-protein and protein-lipid membrane interactions, which should encourage more in-depth future studies. Finally, the development of production and purification methods for recombinant membrane protein p22phox, particularly in native nanodiscs, adds new perspectives to this work that blends experimental and theoretical approaches
3
Launay, Romain. "Computational characterization and understanding of protein assemblies : the case of the Escherichia coli Ubi metabolon involved in ubiquinone biosynthesis." Electronic Thesis or Diss., Toulouse, INSA, 2023. http://www.theses.fr/2023ISAT0055.
Анотація:
Les interactions protéine-protéine (PPIs) et les assemblages supramoléculaires sont essentiels pour les fonctions des cellules vivantes. Ils jouent un rôle important dans un certain nombre de fonctions biologiques, comme la transduction de signaux, la communication entre cellules, la transcription, la réplication ou le transport membranaire. La détermination et la caractérisation de telles interfaces restent un défi en biologie structurale. Cependant, les progrès dans le développement de méthodes computationnelles et la puissance des ressources informatiques disponibles de nos jours ont permis une amélioration considérable de la précision des prédictions in silico des modèles tri-dimensionnels d’assemblages protéiques.Dans le cadre de cette thèse, l’objectif était de prédire la structure d’un assemblage supramoléculaire, appelé métabolon Ubi, impliqué dans la voie de biosynthèse de l’ubiquinone (UQ8) dans Escherichia coli. L’ubiquinone est un prénol possédant des propriétés oxydo-réductrices, localisé dans les membranes, et très conservé à travers l’évolution mais également dans les différentes cellules des organismes. Elle est composée de deux parties principales, un groupe aromatique aux propriétés oxydo-réductrices, appelé quinone ou tête polaire, et une queue polyisoprénoide qui est de nature hydrophobe. Dans le cadre de cette étude, ce sont les dernières étapes de la voie de biosynthèse, notamment les modifications (méthylations et hydroxylations) de la tête polaire, qui nous intéressent. Ces réactions ont lieu au sein du métabolon Ubi. Ce dernier est constitué de sept protéines différentes (UbiE, UbiG, UbiF, UbiH, UbiI, UbiJ, UbiK) catalysant six réactions enzymatiques consécutives.Dans ce travail, nous avons cherché à prédire la structure du métabolon et nous avons ainsi été capable de proposer un sous-ensemble protéique que nous avons nommé "sous-unité centrale". Cette sous-unité comprend l’ensemble des partenaires et pourrait être biologiquement fonctionnelle. En parallèle, une étude a été menée sur l’hétérotrimère UbiJ-UbiK2, une brique moléculaire essentielle du métabolon Ubi. Un modèle tri-dimensionnel de UbiJ-UbiK2 a été proposé. A l’aide d’une étude par modélisation multi-échelles, il a pu être montré qu’il pouvait être impliqué dans le relargage de l’ubiquinone au sein des membranes. Enfin, la dernière partie de ce travail a porté sur l’étude du comportement d’une famille particulière d’enzymes, les mono-oxygénases à flavine de classe A, à laquelle appartiennent UbiF, UbiH et UbiI. Une étude comparative entre une enzyme modèle de cette famille enzymatique, appelée PHBH, et UbiI a été réalisée et concluant à la nécessité d’interactions avec des partenaires, permettant de stabiliser ces protéines au sein du métabolon Ubi.L’ensemble de ces travaux, et des hypothèses proposées, permet d’apporter un regard nouveau sur l’organisation supramoléculaire du métabolon Ubi, tant au niveau structural que fonctionnel. Ainsi, nos résultats ouvrent de nouvelles perspectives pour leur étude expérimentaleProtein-protein interactions (PPIs) and supramolecular assemblies are essential for the functions of living cells. They play an important role in various biological functions, such as signal transduction, cell-cell communication, transcription, replication and membrane transport. Determining and characterizing such interfaces remains a challenge in structural biology. However, advances in the development of computational methods and the power of the computing resources available today have led to a considerable improvement in the accuracy of in silico predictions of three-dimensional models of protein assemblies.In this thesis, the aim was to predict the structure of a supramolecular assembly, called the Ubi metabolon, involved in the ubiquinone (UQ8) biosynthesis pathway in Escherichia coli. Ubiquinone is a prenol with oxido-reducing properties, localized in membranes, and highly conserved throughout evolution but also in different cells of organisms. It is composed of two main parts, an aromatic group with oxido-reducing properties, known as quinone or polar head, and a polyisoprenoid tail which is hydrophobic in nature. In this study, we are interested in the final stages of the biosynthetic pathway, in particular the modifications (methylations and hydroxylations) of the polar head. These reactions take place within the Ubi metabolon. The latter is made up of seven different proteins (UbiE, UbiG, UbiF, UbiH, UbiI, UbiJ, UbiK) catalysing six consecutive enzymatic reactions.In this work, we sought to predict the structure of the metabolon and were thus able to propose a protein subset that we called the 'core subunit'. This sub-unit includes all the partners and could be biologically functional. In parallel, a study was carried out on the UbiJ-UbiK2 heterotrimer, an essential molecular brick of the Ubi metabolon. A three-dimensional model of UbiJ-UbiK2 was proposed. Using a multi-scale modelling study, it was shown that it could be involved in the release of ubiquinone from membranes. Finally, the last part of this work focused on studying the behavior of a particular family of enzymes, the class A flavin mono-oxygenases, to which UbiF, UbiH and UbiI belong. A comparative study between a representative enzyme from this family, called PHBH, and UbiI was carried out, concluding that interactions with partners were necessary to stabilize these proteins within the Ubi metabolon.Taken together, this work and the proposed hypotheses provide a new insight into the supramolecular organization of the Ubi metabolon, both structurally and functionally. Our results open up new prospects for their experimental study
4
Bruley, Apolline. "Exploitation de signatures des repliements protéiques pour décrire le continuum ordre/désordre au sein des protéomes." Electronic Thesis or Diss., Sorbonne université, 2022. http://www.theses.fr/2022SORUS474.
Анотація:
Une fraction significative des protéomes reste non annotée, laissant inaccessible une partie du répertoire fonctionnel de la vie, incluant des innovations moléculaires ayant une valeur thérapeutique ou environnementale. Le manque d'annotation fonctionnelle est en partie dû aux limites des approches actuelles pour la détection de relations cachées, ou à des caractéristiques spécifiques telles que le désordre. L'objectif de ma thèse a été de développer des approches méthodologiques reposant sur les signatures structurales des domaines repliés, afin de caractériser plus avant les séquences protéiques dont la fonction est inconnue, même en l'absence d'informations évolutives. Tout d'abord, j'ai développé un score permettant d'estimer le potentiel de repliement d'une séquence d'acides aminés, basé sur sa densité en amas hydrophobes, correspondant principalement aux structures secondaires régulières. J'ai décrit le continuum entre l'ordre et le désordre, couvrant différents états allant des conformations étendues aux globules fondus et ai caractérisé des cas d'ordre conditionnel. Ensuite, j'ai combiné ce score avec les prédictions de structure 3D d'AlphaFold2 (AF2) disponibles pour 21 protéomes de référence. Une grande fraction des acides aminés des modèles AF2 associés à un très faible index de confiance est incluse dans des segments non repliables, soutenant la qualité d'AF2 comme prédicteur du désordre. Cependant, dans chaque protéome, de longs segments repliables avec des prédictions AF2 de faible confiance présentent également des caractéristiques de domaines solubles et repliés. Cela suggère un ordre caché (conditionnel ou inconditionnel), qui n'est pas détecté par AF2 en raison du manque d'informations évolutives, ou des motifs de repliement non répertoriés. Enfin, à l'aide de ces outils, j'ai effectué une exploration préliminaire de protéines ou de régions non annotées, identifiées via le développement et l'application d'une nouvelle procédure d'annotation. Bien que ces séquences soient enrichies en désordre, une part importante d'entre elles présente des caractéristiques de type globulaire soluble. Ces séquences constituent de bons candidats pour de futures validations et caractérisations expérimentales. De plus, l'analyse de gènes de novo validés expérimentalement m'a permis de contribuer au débat encore ouvert sur les caractéristiques structurales des protéines codées par ces gènes, qui présentent un enrichissement en désordre et une grande diversité d'états structurauxA significant fraction of the proteomes remains unannotated, leaving inaccessible a part of the functional repertoire of life, including molecular innovations with therapeutic or environmental value. Lack of functional annotation is partly due to the limitations of the current approaches in detecting hidden relationships, or to specific features such as disorder. The aim of my PhD thesis was to develop methodological approaches based on the structural signatures of folded domains, in order to further characterize the protein sequences with unknown function even in absence of evolutionary information. First, I developed a scoring system in order to estimate the foldability potential of an amino acid sequence, based on its density in hydrophobic clusters, which mainly correspond to regular secondary structures. I disentangled the continuum between order and disorder, covering various states from extended conformations (random coils) to molten globules and characterize cases of conditional order. Next, I combined this scoring system with the AlphaFold2 (AF2) 3D structure predictions available for 21 reference proteomes. A large fraction of the amino acids with very low AF2 model confidence are included in non-foldable segments, supporting the quality of AF2 as a predictor of disorder. However, within each proteome, long segments with very low AF2 model confidence also exhibit characteristics of soluble, folded domains. This suggests hidden order (conditional or unconditional), which is undetected by AF2 due to lack of evolutionary information, or unrecorded folding patterns. Finally, using these tools, I made a preliminary exploration of unannotated proteins or regions, identified through the development and application of a new annotation workflow. Even though these sequences are enriched in disorder, an important part of them showcases soluble globular-like characteristics. These would make good candidates for further experimental validation and characterization. Moreover, the analysis of experimentally validated de novo genes allowed me to contribute to the still-open debate on the structural features of proteins encoded by these genes, enriched in disorder and displaying a great diversity of structura
Частини книг з теми "AlphaFold":
1
Laurent, Nussaume, Desnos Thierry, Jinsheng Zhu, David Pascale, Kumiko Miwa, and Kanno Satomi. "Analysis and Comparison of Alphafold-Structure Predictions between Pi-Uptake Transporters Recovering Phosphate in Natural Environments." In Plant Phosphorus Nutrition, 129–50. Boca Raton: CRC Press, 2023. http://dx.doi.org/10.1201/9781003440079-9.
2
Moriarty, David. "Alphafold 2." In Designing and Managing Complex Systems, 263–66. Elsevier, 2023. http://dx.doi.org/10.1016/b978-0-323-91609-7.00026-3.
3
Mariano, Diego. "AlphaFold e a busca pelo Santo Graal da Biologia Molecular." In BIOINFO #02 - Revista Brasileira de Bioinformática e Biologia Computacional, 162–67. 2nd ed. Alfahelix, 2022. http://dx.doi.org/10.51780/978-65-992753-5-7-10.
Анотація:
Recentemente, o uso de técnicas de inteligência artificial tem proporcionado um grande avanço para a solução de um problema da biologia estrutural estudado há mais de 50 anos, trazendo novas perspectivas para a compreensão de doenças e para a descoberta de novos medicamentos.
4
Jones, David T., and Janet M. Thornton. "AlphaFold — The End of the Protein Folding Problem or the Start of Something Bigger?" In Artificial Intelligence for Science, 67–80. WORLD SCIENTIFIC, 2023. http://dx.doi.org/10.1142/9789811265679_0005.
5
Blaney, Jeff, and Andrew M. Davis. "Structure-based Design for Medicinal Chemists." In The Handbook of Medicinal Chemistry, 137–87. The Royal Society of Chemistry, 2023. http://dx.doi.org/10.1039/9781788018982-00137.
Анотація:
Structure-based design is an essential part of medicinal chemistry. The availability of experimental structures for many drug discovery targets and improvements in molecular modeling software makes it practical for medicinal chemists to do their own modeling and design. We will focus on how we've seen structure-based design performed during many medicinal chemistry projects in industry. Structure-based design relies most critically on a solid understanding of physical organic chemistry, especially conformational analysis and intra- and intermolecular interactions, and these aspects are covered in depth together with illustrative case studies. The amazing rapid advances in cryo-EM over the last several years are the most exciting and dramatic advance for structure-based design since the previous edition of this book. Those advances have enabled structure determination of membrane proteins and large protein complexes. The breakthroughs in protein structure prediction made by AlphaFold and RoseTTAFold are also considered. Future directions for application of structure-based drug design are considered including use of machine learning and large-scale virtual screening, and PROTAC and molecular glue design.
6
Ganguly, Rik, Shashi Kumar Yadav, Prosperwell Ingty, Angneh Ngoruh, and Atanu Bhattacharjee. "LEVERAGING THE STRENGTH OF ARTIFICIAL INTELLIGENCE IN SOLVING PROTEIN STRUCTURES BY ALPHAFOLD-2- A MODERN APPROACH TO UNDERSTAND PROTEIN DYNAMICS." In Futuristic Trends in Biotechnology Volume 2 Book 27, 234–45. Iterative International Publishers, Selfypage Developers Pvt Ltd, 2023. http://dx.doi.org/10.58532/v2bs27p3ch1.
Анотація:
Proteins are unique macromolecules made up of a long chain of amino acids and are classified based on their function, structure, shape, chemical composition and solubility in different solvents. A wide variety of proteins are prone to misfold and create intracellular or extracellular aggregates that cause severe cellular malfunction. The importance of the protein folding problem was recognized and put forward 50 years back by distinguished scientists. Understanding the dynamics of protein folding is crucial and this can help us predict the ultimate configuration of functional protein. Many of the life-threatening diseases are caused by the misfolding of proteins. The reason for the misfolding can be point mutations since the three-dimensional structure of proteins depends on the primary sequence of its amino acid. Despite fifty years of research, we still need to fill the knowledge gap and accelerate our understanding, particularly in computational biology for the accurate prediction of protein structure. Homology modeling is utilized to predict protein structure in absence of experimental structure. Artificial intelligence, machine learning, and deep learning are being used extensively by researchers to computationally estimate a protein's structure based only on its amino acid sequence. AlphaFold which is in a second iteration tool has changed the perception about protein folding by solving the unsolved structures
7
Muhammed, Muhammed Tilahun, and Esin Aki-Yalcin. "Up-to-Date Developments in Homology Modeling." In Applied Computer-Aided Drug Design: Models and Methods, 116–35. BENTHAM SCIENCE PUBLISHERS, 2023. http://dx.doi.org/10.2174/9789815179934123010006.
Анотація:
Homology modeling is used to predict protein 3D structure from its amino acid sequence. It is the most accurate computational approach to estimate 3D structures. It has straightforward steps that save time and labor. There are several homology modeling tools under use. There is no sole tool that is superior in every aspect. Hence, the user should select the most appropriate one carefully. It is also a common practice to use two or more tools at a time and choose the best model among the resulting models. Homology modeling has various applications in the drug design and development process. Such applications need high-quality 3D structures. It is widely used in combination with other computational methods including molecular docking and molecular dynamics simulation. Like the other computational methods, it has been influenced by the involvement of artificial intelligence. In this regard, homology modeling tools, like AlphaFold, have been introduced. This type of method is expected to contribute to filling the gap between protein sequence release and 3D structure determination. This chapter sheds light on the history, relatively popular tools and steps of homology modeling. A detailed explanation of MODELLER is also given as a case study protocol. Furthermore, homology modeling’s application in drug discovery is explained by exemplifying its role in the fight against the novel Coronavirus. Considering the new advances in the area, better tools and thus high-quality models are expected. These, in turn, pave the way for more applications of it.
8
Mubarak, Malad, Aastha Senapati, Purva Bankar, Nirmitee Dolas, Jyoti Srivastava, Shankar Mukundrao Khade, and Krishna Kant Pandey. "Introduction to AI in Biotechnology and Biomedical Engineering." In Future of AI in Biomedicine and Biotechnology, 1–17. IGI Global, 2024. http://dx.doi.org/10.4018/979-8-3693-3629-8.ch001.
Анотація:
Artificial intelligence (AI) is a field of computer science that works towards using machines/computer programs to perform tasks that would normally require human intelligence. Some of these tasks may include prediction, problem-solving, reasoning, and inferring. Some the most notable uses are protein structure prediction (AlphaFold2) and the prediction of disease-causing genetic mutations in primates (PrimateAI-3D). Machine learning models have also found use in diagnosis. A study showed that it is possible to get accurate diagnosis from limited amounts of data if the data is handled and processed properly. Overall, AI can be applied in a wide range of fields, given that accurate and sufficient data is provided, and this is just what is known. There are many more applications that might come up over time. This chapter aims to give a basic overview of machine learning algorithms used to train models and shed some light on how it is being used in the fields of biotechnology and biomedical engineering.
Тези доповідей конференцій з теми "AlphaFold":
1
Maghiar, Octavian-Florin. "AlphaFold-based protein analysis pipeline." In 2022 24th International Symposium on Symbolic and Numeric Algorithms for Scientific Computing (SYNASC). IEEE, 2022. http://dx.doi.org/10.1109/synasc57785.2022.00061.
2
Смирнов, Сергей Владимирович. "SOLVING CURRENT PROBLEMS USING NEURAL NETWORKS." In Современные методы и инновации в науке: сборник статей XXIII международной научной конференции (Санкт-Петербург, Декабрь 2023). Crossref, 2024. http://dx.doi.org/10.37539/231208.2023.38.36.003.
Анотація:
Работа посвящена применению нейронных сетей, которые успешно решают самые актуальные проблемы, стоящие перед современным обществом. Представлена краткая история развития нейросетей. Рассматриваются следующие нейронные сети: ChatGPT, MENNDL, AlphaFold, Exscientia, Neuralangelo, WomboArt и ModelScope. Обрисованы возможности дальнейшего развития нейронных сетей. The work is devoted to the use of neural networks, which successfully solve the most pressing problems facing modern society. Представлена краткая история развития нейросетей. The following neural networks are considered: ChatGPT, MENNDL, AlphaFold, Exscientia, Neuralangelo, WomboArt и ModelScope. Possibilities for further development of neural networks are outlined.
3
Tan, Juntao, and Yongfeng Zhang. "ExplainableFold: Understanding AlphaFold Prediction with Explainable AI." In KDD '23: The 29th ACM SIGKDD Conference on Knowledge Discovery and Data Mining. New York, NY, USA: ACM, 2023. http://dx.doi.org/10.1145/3580305.3599337.
4
Zhong, Bozitao, Xiaoming Su, Minhua Wen, Sicheng Zuo, Liang Hong, and James Lin. "ParaFold: Paralleling AlphaFold for Large-Scale Predictions." In HPCAsia 2022 Workshop: International Conference on High Performance Computing in Asia-Pacific Region Workshops. New York, NY, USA: ACM, 2022. http://dx.doi.org/10.1145/3503470.3503471.
5
Chowdhury, Abu Sayeed, and Julia Thom Oxford. "Collagen a1 (XI) structure prediction by Alphafold 2." In 2022 International Conference on Computational Science and Computational Intelligence (CSCI). IEEE, 2022. http://dx.doi.org/10.1109/csci58124.2022.00108.
6
Srivastava, Pragya, Shreyansh Suyash, and N. Jayapandian. "Artificial Intelligence based System in Protein Folding using Alphafold." In 2022 International Conference on Automation, Computing and Renewable Systems (ICACRS). IEEE, 2022. http://dx.doi.org/10.1109/icacrs55517.2022.10029190.
7
Cheng, Shenggan, Xuanlei Zhao, Guangyang Lu, Jiarui Fang, Tian Zheng, Ruidong Wu, Xiwen Zhang, Jian Peng, and Yang You. "FastFold: Optimizing AlphaFold Training and Inference on GPU Clusters." In PPoPP '24: 29th ACM SIGPLAN Annual Symposium on Principles and Practice of Parallel Programming. New York, NY, USA: ACM, 2024. http://dx.doi.org/10.1145/3627535.3638465.
8
Fujita, Hayato, Akihiro Nomura, Toshio Endo, and Masakazu Sekijima. "Enhancing the Performance of AlphaFold Through Modified Storage Method and Optimization of HHblits on TSUBAME3.0 Supercomputer." In 2023 Congress in Computer Science, Computer Engineering, & Applied Computing (CSCE). IEEE, 2023. http://dx.doi.org/10.1109/csce60160.2023.00351.
9
Sánchez-Galan, Javier. "MACHINE LEARNING Y SUS APLICACIONES." In V Congreso de Investigación Desarrollo en Innovación de la Universidad Internacional de Ciencia y Tecnología. Universidad Internacional de Ciencia y Tecnología, 2021. http://dx.doi.org/10.47300/978-9962-5599-8-6-03.
Анотація:
En la charla se presentaron conceptos básicos e historia de la metodología de desarrollo de software llamada Machine Learning (ML) o aprendizaje automático (AA, en español. Se explicaron los tres distintos tipos de aprendizaje que existen dentro del AA, aprendizaje supervisado, no-supervisado y por refuerzo; que se utilizan dependiendo de la tarea que se quiere lograr. Se expusieron las diversas aplicaciones que ha abordado el grupo de investigación utilizando esta metodología para preguntas de investigación de relevancia nacional. Desde ayuda en análisis de bioinformática (apoyando a la determinación espectral-molecular de especies de mosquitos), ecología (determinación y clasificación de vocalizaciones de manatíes) y aplicaciones en agricultura (para la determinación automática de variedades de plantas arroz y para la clasificación automática de sandias para exportación). La charla concluyó con una reflexión de los últimos avances de esta metodología y se presentó el reciente caso de la predicción del doblaje en proteínas con el algoritmo Alphafold 2.
10
Konc, Janez, and Dušanka Janežič. "Algorithms and web servers for protein binding sites detection in drug discovery." In 2nd International Conference on Chemo and BioInformatics. Institute for Information Technologies, University of Kragujevac, 2023. http://dx.doi.org/10.46793/iccbi23.014k.
Анотація:
Drug discovery is a protracted and demanding process, which can be expedited during its early stages through novel mathematical approaches and modern computing. To tackle this crucial issue, we are developing fresh mathematical solutions aimed at detecting and characterizing protein binding sites, pivotal for new drug discovery. This paper introduces algorithms founded on graph theory which we have devised to scrutinize target biological proteins. These algorithms yield vital data, facilitating the optimization of initial phases in novel drug development. A particular emphasis lies in the creation of pioneering protein binding site prediction algorithms (ProBiS) and innovative web tools for modeling pharmaceutically intriguing molecules—ProBiS tools. These tools have matured into comprehensive graphical resources for the study of proteins in the proteome. ProBiS stands apart from other structural algorithms due to its ability to align proteins with disparate folds, all without prior knowledge of the binding sites. This unique capability enables the identification of analogous binding sites and the prediction of molecular ligands of diverse pharmaceutical relevance. These ligands could potentially progress into drug candidates for treating diseases. Notably, this prediction is based on data sourced from the complete Protein Data Bank (PDB) and the AlphaFold database, encompassing proteins not yet cataloged in the PDB. All ProBiS tools are made available without charge to the academic community through http://insilab.org and https://probis.nih.gov.