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

---

The unintended consequences of COVID-19 vaccine policy: why mandates, passports and restrictions may cause more harm than good

BMJ Global Health, Journal Year: 2022, Volume and Issue: 7(5), P. e008684 - e008684

Published: May 1, 2022

Vaccination policies have shifted dramatically during COVID-19 with the rapid emergence of population-wide vaccine mandates, domestic passports and differential restrictions based on vaccination status. While these prompted ethical, scientific, practical, legal political debate, there has been limited evaluation their potential unintended consequences. Here, we outline a comprehensive set hypotheses for why may ultimately be counterproductive harmful. Our framework considers four domains: (1) behavioural psychology, (2) politics law, (3) socioeconomics, (4) integrity science public health. current vaccines appear to had significant impact decreasing COVID-19-related morbidity mortality burdens, argue that mandatory are scientifically questionable likely cause more societal harm than good. Restricting people’s access work, education, transport social life status impinges human rights, promotes stigma polarisation, adversely affects health well-being. Current lead widening economic inequalities, detrimental long-term impacts trust in government scientific institutions, reduce uptake future measures, including as well routine immunisations. Mandating is one most powerful interventions should used sparingly carefully uphold ethical norms institutions. We re-evaluated light negative consequences outline. Leveraging empowering strategies consultation, improving healthcare services infrastructure, represent sustainable approach optimising programmes and, broadly, well-being public.

Language: Английский

---

Using AlphaFold to predict the impact of single mutations on protein stability and function

PLoS ONE, Journal Year: 2023, Volume and Issue: 18(3), P. e0282689 - e0282689

Published: March 16, 2023

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 folding problem is "solved". However, more than just sequence. Presently, it unknown if AlphaFold-triggered revolution could help solve other problems related folding. Here we assay ability predict impact single mutations on stability (ΔΔG) and function. To study question extracted pLDDT metrics predictions before after mutation in a correlated predicted change with experimentally known ΔΔG values. Additionally, same using large scale dataset GFP assayed levels fluorescence. We found very weak or no correlation between output Our results imply may not be immediately applied applications

Language: Английский

---

Metapredict: a fast, accurate, and easy-to-use predictor of consensus disorder and structure

Biophysical Journal, Journal Year: 2021, Volume and Issue: 120(20), P. 4312 - 4319

Published: Sept. 2, 2021

Language: Английский

---

OpenFold: Retraining AlphaFold2 yields new insights into its learning mechanisms and capacity for generalization

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

Published: Nov. 22, 2022

Abstract AlphaFold2 revolutionized structural biology with the ability to predict protein structures exceptionally high accuracy. Its implementation, however, lacks code and data required train new models. These are necessary (i) tackle tasks, like protein-ligand complex structure prediction, (ii) investigate process by which model learns, remains poorly understood, (iii) assess model’s generalization capacity unseen regions of fold space. Here we report OpenFold, a fast, memory-efficient, trainable implementation AlphaFold2. We OpenFold from scratch, fully matching accuracy Having established parity, OpenFold’s generalize across space retraining it using carefully designed datasets. find that is remarkably robust at generalizing despite extreme reductions in training set size diversity, including near-complete elisions classes secondary elements. By analyzing intermediate produced during training, also gain surprising insights into manner learns proteins, discovering spatial dimensions learned sequentially. Taken together, our studies demonstrate power utility believe will prove be crucial resource for modeling community.

Language: Английский

---

OpenFold: retraining AlphaFold2 yields new insights into its learning mechanisms and capacity for generalization

Nature Methods, Journal Year: 2024, Volume and Issue: 21(8), P. 1514 - 1524

Published: May 14, 2024

AlphaFold2 revolutionized structural biology with the ability to predict protein structures exceptionally high accuracy. Its implementation, however, lacks code and data required train new models. These are necessary (1) tackle tasks, like protein–ligand complex structure prediction, (2) investigate process by which model learns (3) assess model's capacity generalize unseen regions of fold space. Here we report OpenFold, a fast, memory efficient trainable implementation AlphaFold2. We OpenFold from scratch, matching accuracy Having established parity, find that is remarkably robust at generalizing even when size diversity its training set deliberately limited, including near-complete elisions classes secondary elements. By analyzing intermediate produced during training, also gain insights into hierarchical manner in fold. In sum, our studies demonstrate power utility believe will prove be crucial resource for modeling community. open-source It fast efficient, available under permissive license.

Language: Английский

---

Ins and outs of AlphaFold2 transmembrane protein structure predictions

Cellular and Molecular Life Sciences, Journal Year: 2022, Volume and Issue: 79(1)

Published: Jan. 1, 2022

Transmembrane (TM) proteins are major drug targets, but their structure determination, a prerequisite for rational design, remains challenging. Recently, the DeepMind's AlphaFold2 machine learning method greatly expanded structural coverage of sequences with high accuracy. Since employed algorithm did not take specific properties TM into account, reliability generated structures should be assessed. Therefore, we quantitatively investigated quality at genome scales, level ABC protein superfamily folds and membrane (e.g. dimer modeling stability in molecular dynamics simulations). We tested template-free prediction challenging CASP14 target several published after training. Our results suggest that performs well case its neural network is overfitted. conclude cautious applications models will advance protein-associated studies an unexpected level.

Language: Английский

---

Improving peptide-protein docking with AlphaFold-Multimer using forced sampling

Frontiers in Bioinformatics, Journal Year: 2022, Volume and Issue: 2

Published: Sept. 26, 2022

Protein interactions are key in vital biological processes. In many cases, particularly regulation, this interaction is between a protein and shorter peptide fragment. Such peptides often part of larger disordered regions other proteins. The flexible nature enables the rapid yet specific regulation important functions cells, such as their life cycle. Consequently, knowledge molecular details peptide-protein crucial for understanding altering function, specialized computational methods have been developed to study them. recent release AlphaFold AlphaFold-Multimer has led leap accuracy modeling study, ability predict which proteins interact, well its resulting complexes, benchmarked against established methods. We find that predicts structure complexes with acceptable or better quality (DockQ ≥0.23) 66 112 investigated—25 were high ≥0.8). This massive improvement on previous 23 47 models only four eight models, when using energy-based docking templates, respectively. addition, can be used whether will interact. At 1% false positives, found 26% possible precision 85%, best among benchmarked. However, most interesting result possibility improving by randomly perturbing neural network weights force sample more conformational space. increases number from 75 improves median DockQ 0.47 0.55 (17%) first ranked models. 0.58 0.72 (24%), indicating selecting model still challenge. scheme generating structures should generally useful applications involving multiple states, regions, disorder.

Language: Английский

---

The HADDOCK2.4 web server for integrative modeling of biomolecular complexes

Nature Protocols, Journal Year: 2024, Volume and Issue: unknown

Published: June 17, 2024

Language: Английский

---

Benchmarking AlphaFold2 on peptide structure prediction

Structure, Journal Year: 2022, Volume and Issue: 31(1), P. 111 - 119.e2

Published: Dec. 15, 2022

Language: Английский

---