AlphaFold2 and its applications in the fields of biology and medicine

Signal Transduction and Targeted Therapy, Journal Year: 2023, Volume and Issue: 8(1)

Published: March 14, 2023

Abstract AlphaFold2 (AF2) is an artificial intelligence (AI) system developed by DeepMind that can predict three-dimensional (3D) structures of proteins from amino acid sequences with atomic-level accuracy. Protein structure prediction one the most challenging problems in computational biology and chemistry, has puzzled scientists for 50 years. The advent AF2 presents unprecedented progress protein attracted much attention. Subsequent release more than 200 million predicted further aroused great enthusiasm science community, especially fields medicine. thought to have a significant impact on structural research areas need information, such as drug discovery, design, function, et al. Though time not long since was developed, there are already quite few application studies medicine, many them having preliminarily proved potential AF2. To better understand promote its applications, we will this article summarize principle architecture well recipe success, particularly focus reviewing applications Limitations current also be discussed.

Language: Английский

---

Predicting multiple conformations via sequence clustering and AlphaFold2

Nature, Journal Year: 2023, Volume and Issue: 625(7996), P. 832 - 839

Published: Nov. 13, 2023

AlphaFold2 (ref. 1) has revolutionized structural biology by accurately predicting single structures of proteins. However, a protein's biological function often depends on multiple conformational substates2, and disease-causing point mutations cause population changes within these substates3,4. We demonstrate that clustering multiple-sequence alignment sequence similarity enables to sample alternative states known metamorphic proteins with high confidence. Using this method, named AF-Cluster, we investigated the evolutionary distribution predicted for protein KaiB5 found predictions both conformations were distributed in clusters across KaiB family. used nuclear magnetic resonance spectroscopy confirm an AF-Cluster prediction: cyanobacteria variant is stabilized opposite state compared more widely studied variant. To test AF-Cluster's sensitivity mutations, designed experimentally verified set three flip from Rhodobacter sphaeroides ground fold-switched state. Finally, screening families without fold switching identified putative oxidoreductase Mpt53 Mycobacterium tuberculosis. Further development such bioinformatic methods tandem experiments will probably have considerable impact energy landscapes, essential illuminating function.

Language: Английский

---

SPEACH_AF: Sampling protein ensembles and conformational heterogeneity with Alphafold2

PLoS Computational Biology, Journal Year: 2022, Volume and Issue: 18(8), P. e1010483 - e1010483

Published: Aug. 22, 2022

The unprecedented performance of Deepmind's Alphafold2 in predicting protein structure CASP XIV and the creation a database structures for multiple proteomes sequence repositories is reshaping structural biology. However, because this returns single structure, it brought into question Alphafold's ability to capture intrinsic conformational flexibility proteins. Here we present general approach drive model alternate conformations through simple manipulation alignment via silico mutagenesis. grounded hypothesis that must also encode heterogeneity, thus its rational will enable sample conformations. A systematic modeling pipeline benchmarked against canonical examples applied interrogate landscape membrane This work broadens applicability by generating be tested biologically, biochemically, biophysically, use structure-based drug design.

Language: Английский

---

AlphaFold2 fails to predict protein fold switching

Protein Science, Journal Year: 2022, Volume and Issue: 31(6)

Published: May 26, 2022

AlphaFold2 has revolutionized protein structure prediction by leveraging sequence information to rapidly model folds with atomic-level accuracy. Nevertheless, previous work shown that these predictions tend be inaccurate for structurally heterogeneous proteins. To systematically assess factors contribute this inaccuracy, we tested AlphaFold2's performance on 98-fold-switching proteins, which assume at least two distinct-yet-stable secondary and tertiary structures. Topological similarities were quantified between five predicted experimentally determined structures of each fold-switching protein. Overall, 94% captured one conformation but not the other. Despite biased results, estimated confidences moderate-to-high 74% residues, a result contrasts overall low intrinsically disordered are also heterogeneous. investigate contributing disparity, variation within multiple alignments used generate Unlike regions, whose show conservation, regions had conservation rates statistically similar canonical single-fold Furthermore, lower than either or regardless conservation. high fold switchers indicate it uses sophisticated pattern recognition search most probable conformer rather biophysics protein's structural ensemble. Thus, is surprising its often fail proteins properties fully apparent from solved Our results emphasize need look as an ensemble suggest systematic examination sequences may reveal propensities stable

Language: Английский

---

Protein structure prediction has reached the single-structure frontier

Nature Methods, Journal Year: 2023, Volume and Issue: 20(2), P. 170 - 173

Published: Jan. 13, 2023

Language: Английский

---

High-throughput prediction of protein conformational distributions with subsampled AlphaFold2

Nature Communications, Journal Year: 2024, Volume and Issue: 15(1)

Published: March 27, 2024

Abstract This paper presents an innovative approach for predicting the relative populations of protein conformations using AlphaFold 2, AI-powered method that has revolutionized biology by enabling accurate prediction structures. While 2 shown exceptional accuracy and speed, it is designed to predict proteins’ ground state limited in its ability conformational landscapes. Here, we demonstrate how can directly different subsampling multiple sequence alignments. We tested our against nuclear magnetic resonance experiments on two proteins with drastically amounts available data, Abl1 kinase granulocyte-macrophage colony-stimulating factor, predicted changes their more than 80% accuracy. Our worked best when used qualitatively effects mutations or evolution landscape well-populated states proteins. It thus offers a fast cost-effective way at even single-point mutation resolution, making useful tool pharmacology, analysis experimental results, evolution.

Language: Английский

---

Prediction of multiple conformational states by combining sequence clustering with AlphaFold2

bioRxiv (Cold Spring Harbor Laboratory), Journal Year: 2022, Volume and Issue: unknown

Published: Oct. 17, 2022

Abstract AlphaFold2 (AF2) has revolutionized structural biology by accurately predicting single structures of proteins and protein-protein complexes. However, biological function is rooted in a protein’s ability to sample different conformational substates, disease-causing point mutations are often due population changes these substates. This sparked immense interest expanding AF2’s capability predict We demonstrate that clustering an input multiple sequence alignment (MSA) similarity enables AF2 alternate states known metamorphic proteins, including the circadian rhythm protein KaiB, transcription factor RfaH, spindle checkpoint Mad2, score with high confidence. Moreover, we use identify minimal set two predicted switch KaiB between its states. Finally, used our method, AF-cluster, screen for families without fold-switching, identified putative state oxidoreductase DsbE. Similarly DsbE thioredoxin-like fold novel fold. prediction subject future experimental testing. Further development such bioinformatic methods tandem experiments will likely have profound impact on energy landscapes, essential shedding light into function.

Language: Английский

---

Protein structure generation via folding diffusion

Nature Communications, Journal Year: 2024, Volume and Issue: 15(1)

Published: Feb. 5, 2024

Abstract The ability to computationally generate novel yet physically foldable protein structures could lead new biological discoveries and treatments targeting incurable diseases. Despite recent advances in structure prediction, directly generating diverse, from neural networks remains difficult. In this work, we present a diffusion-based generative model that generates backbone via procedure inspired by the natural folding process. We describe as sequence of angles capturing relative orientation constituent atoms, denoising random, unfolded state towards stable folded structure. Not only does mirror how proteins natively twist into energetically favorable conformations, inherent shift rotational invariance representation crucially alleviates need for more complex equivariant networks. train diffusion probabilistic with simple transformer demonstrate our resulting unconditionally highly realistic complexity structural patterns akin those naturally-occurring proteins. As useful resource, release an open-source codebase trained models diffusion.

Language: Английский

---

AlphaFold predictions of fold-switched conformations are driven by structure memorization

Nature Communications, Journal Year: 2024, Volume and Issue: 15(1)

Published: Aug. 24, 2024

Abstract Recent work suggests that AlphaFold (AF)–a deep learning-based model can accurately infer protein structure from sequence–may discern important features of folded energy landscapes, defined by the diversity and frequency different conformations in state. Here, we test limits its predictive power on fold-switching proteins, which assume two structures with regions distinct secondary and/or tertiary structure. We find (1) AF is a weak predictor fold switching (2) some successes result memorization training-set rather than learned energetics. Combining >280,000 models several implementations AF2 AF3, 35% success rate was achieved for switchers likely AF’s training sets. AF2’s confidence metrics selected against consistent experimentally determined failed to discriminate between low high conformations. Further, captured only one out seven confirmed outside sets despite extensive sampling an additional ~280,000 models. Several observations indicate has memorized structural information during training, AF3 misassigns coevolutionary restraints. These limitations constrain scope successful predictions, highlighting need physically based methods readily predict multiple

Language: Английский

---

The power and pitfalls of AlphaFold2 for structure prediction beyond rigid globular proteins

Nature Chemical Biology, Journal Year: 2024, Volume and Issue: 20(8), P. 950 - 959

Published: June 21, 2024

Language: Английский

---