PubHawk
August 2021 PMID 34293799

Highly accurate protein structure prediction for the human proteome.

Kathryn Tunyasuvunakool, Jonas Adler, Zachary Wu, Tim Green, Michal Zielinski, Augustin Žídek, Alex Bridgland, Andrew Cowie, Clemens Meyer, Agata Laydon, Sameer Velankar, Gerard J Kleywegt, Alex Bateman, Richard Evans, Alexander Pritzel

Full text PDF Download social card

Mentions

165
Peak: 10 in Jul 2021
across 6 sources since Jul 2021

Extended reach

119,502
potential audience

Total likes

190

Video views

6,555

Mentions per month

Jul 2021 – Sep 2025

Where it's discussed

  • News99 articles
  • Blogs30 articles
  • X / Twitter18 posts
  • Facebook8 posts
  • Reddit6 posts
  • YouTube4 videos

AI synopsis

What people are saying

Discussions emphasize the groundbreaking nature of using AlphaFold for the human proteome, with comments like 'very much looking forward' and phrases like 'inspiring paper' signaling strong interest. Several posts critique and analyze the methodology, such as those referencing pLDDT scores as predictors of disordered regions, demonstrating active engagement and recognition of the research's broad impact on structural bioinformatics.

AgreementStrong agreement

Most discussions express strong support, highlighting the significance and promising aspects of the publication.

InterestHigh interest

Participants show high interest, mentioning specific metrics, potential applications, and enthusiasm for the resource.

EngagementHigh engagement

The responses feature detailed analyses, referencing additional studies, and actively discussing implications, indicating deep engagement.

ImpactHigh impact

There is a consensus that this work can significantly advance biological research, drug discovery, and understanding of protein structures.

Abstract in brief

  • The study uses the advanced AI tool AlphaFold to predict the 3D structures of almost all human proteins, greatly increasing the amount of available structural data from only 17% coverage to about 58% of residues with high confidence.
  • The researchers developed new metrics to evaluate these predictions, identifying well-structured regions, multi-domain proteins, and disordered areas, making the data more useful for biological research and drug development.
  • They provide case studies to show how these high-quality models can help generate new biological hypotheses, and they are sharing their data openly to support widespread scientific use.]
Read the full abstract

Top posts

Most liked on X and Bluesky
Titiat Scriptor
@titiatscriptor · X

13| Quellen: Q1: https://t.co/f6TsxM4xkZ Q2: https://t.co/b905GDgtFw Q3: https://t.co/u3C9HexokU Q4: https://t.co/NoERDjhwGK

Jul 12, 2022 · 14 likes · 6,609 followers View post →
Kresten Lindorff-Larsen
@LindorffLarsen · X

@_BalintMeszaros @DaveyLab @DeepMind @emblebi Cool. I assume you have also seen the analysis of pLDDT as a disorder predictor in the new DM/EBI paper (Fig. 4b). https://t.co/wQkXXo5ECQ

Jul 26, 2021 · 8 likes · 14,558 followers View post →
Stephan Hacker (inactive account)
@StephanHacker2 · X

Finally managed to read the #AlphaFold paper by @DeepMind with @emblebi in @Nature. Structural predictions for a large part of the human proteome. Very much looking forward to exploring how this vast data resource can help to guide our research. (1/3) https://t.co/QOYQlPxvnC

Jul 30, 2021 · 7 likes · 5,087 followers View post →
Kresten Lindorff-Larsen
@LindorffLarsen · X

@fabferrage @GntlmnScientist Indeed, low AF confidence scores (pLDDT) is an excellent predictor of intrinsic disorder as shown here for example: https://t.co/wQkXXo5ECQ (Fig. 4b) https://t.co/jlsT72tzDW (Fig. 1d)

Jan 7, 2022 · 6 likes · 1 share · 14,558 followers View post →
Kresten Lindorff-Larsen
@LindorffLarsen · X

@larsjuhljensen DM/EBI themselves showed that pLDDT works well as a disorder predictor (Fig. 4b https://t.co/wQkXXo5ECQ) and this preprint (https://t.co/3iefY5rmfS) uses it to study protein-peptide interactions. A few other threads w similar analysis out there.

Aug 4, 2021 · 6 likes · 14,558 followers View post →
Kresten Lindorff-Larsen
@LindorffLarsen · X

@agrossfield @CrollTristan @boehninglab @UCSFChimeraX @ditlevbrodersen The DM/EBI team showed that the AF2 confidence scores work well as disorder predictors (Fig. 4B of https://t.co/wQkXXo5ECQ) and @_BalintMeszaros has followed up with additional analyses (https://t.co/tXUX3Nujo0), suggesting that many low confidence regions are in fact disordered.

Sep 10, 2021 · 5 likes · 3 shares · 14,558 followers View post →

All mentions

IntegralAnswers
@IntegralAnswers · X

11/

Sep 24, 2025 · 3 likes · 8,309 followers View post →
Kresten Lindorff-Larsen
@LindorffLarsen · X

@GntlmnScientist @ProteinMagnus A low AF pLDDT score is a good predictor of intrinsic disorder; see eg: https://t.co/wQkXXo5ECQ & https://t.co/jlsT72tzDW and disordered regions with high pLDDT scores might be “conditionally folded”: https://t.co/nPoxYy4RHE

Aug 6, 2022 · 1 like · 14,558 followers View post →
Titiat Scriptor
@titiatscriptor · X

13| Quellen: Q1: https://t.co/f6TsxM4xkZ Q2: https://t.co/b905GDgtFw Q3: https://t.co/u3C9HexokU Q4: https://t.co/NoERDjhwGK

Jul 12, 2022 · 14 likes · 6,609 followers View post →
Noelia Ferruz
@ferruz_noelia · X

RT @LindorffLarsen: @fabferrage @GntlmnScientist Indeed, low AF confidence scores (pLDDT) is an excellent predictor of intrinsic disorder a…

Jan 7, 2022 · 1 share · 5,363 followers View post →
Kresten Lindorff-Larsen
@LindorffLarsen · X

@fabferrage @GntlmnScientist Indeed, low AF confidence scores (pLDDT) is an excellent predictor of intrinsic disorder as shown here for example: https://t.co/wQkXXo5ECQ (Fig. 4b) https://t.co/jlsT72tzDW (Fig. 1d)

Jan 7, 2022 · 6 likes · 1 share · 14,558 followers View post →
Tristan Croll
@CrollTristan · X

RT @LindorffLarsen: @agrossfield @CrollTristan @boehninglab @UCSFChimeraX @ditlevbrodersen The DM/EBI team showed that the AF2 confidence s…

Sep 10, 2021 · 3 shares · 3,664 followers View post →
Kresten Lindorff-Larsen
@LindorffLarsen · X

@agrossfield @CrollTristan @boehninglab @UCSFChimeraX @ditlevbrodersen The DM/EBI team showed that the AF2 confidence scores work well as disorder predictors (Fig. 4B of https://t.co/wQkXXo5ECQ) and @_BalintMeszaros has followed up with additional analyses (https://t.co/tXUX3Nujo0), suggesting that many low confidence regions are in fact disordered.

Sep 10, 2021 · 5 likes · 3 shares · 14,558 followers View post →
Kresten Lindorff-Larsen
@LindorffLarsen · X

@larsjuhljensen DM/EBI themselves showed that pLDDT works well as a disorder predictor (Fig. 4b https://t.co/wQkXXo5ECQ) and this preprint (https://t.co/3iefY5rmfS) uses it to study protein-peptide interactions. A few other threads w similar analysis out there.

Aug 4, 2021 · 6 likes · 14,558 followers View post →
Stephan Hacker (inactive account)
@StephanHacker2 · X

Finally managed to read the #AlphaFold paper by @DeepMind with @emblebi in @Nature. Structural predictions for a large part of the human proteome. Very much looking forward to exploring how this vast data resource can help to guide our research. (1/3) https://t.co/QOYQlPxvnC

Jul 30, 2021 · 7 likes · 5,087 followers View post →
Shahryar Alavi
@Schahrjar · X

https://t.co/2Xo33M7CUT “… several metrics … identifying … regions likely to be disordered.” An intrinsically disordered region has no unique structure. They get structured regarding their partners. Thanks @DeepMind for the release of #AlphaFold predicted structures.

Jul 29, 2021 · 61 followers View post →
Hiroki Onoda
@onoda_hiroki · X

メモ: features.pklのaatypeは[残基数, 21]の2Dアレイ、One-hot representation of the input amino acid sequence (20 amino acids + unknown).順番はアミノ酸3文字表記ののアルファベット順のよう https://t.co/8dU8bxdIyl https://t.co/L9gUkoZlC7

Jul 27, 2021 · 531 followers View post →
Kresten Lindorff-Larsen
@LindorffLarsen · X

@_BalintMeszaros @DaveyLab @DeepMind @emblebi Cool. I assume you have also seen the analysis of pLDDT as a disorder predictor in the new DM/EBI paper (Fig. 4b). https://t.co/wQkXXo5ECQ

Jul 26, 2021 · 8 likes · 14,558 followers View post →
Dityatev Lab
@DityatevL · X

Highly accurate protein structure prediction for the human proteome https://t.co/hdMyIuglgP

Jul 24, 2021 · 1 like · 562 followers View post →
Alatorre-Jacome, PhD.
@OscarAlatorre17 · X

https://t.co/etdS3Y1vJY

Jul 24, 2021 · 10 followers View post →
Michal Jasinski
@MedicagoCrew · X

Slightly different but what an inspiring paper! https://t.co/IZxiqwLqME

Jul 23, 2021 · 194 followers View post →
Leon Eyrich Jessen
@jessenleon · X

RT @LindorffLarsen:

Jul 23, 2021 · 2 shares · 1,613 followers View post →
Cesar Ramirez-Sarmiento @cesarramirez.bsky.social
@cxarramirez · X

RT @LindorffLarsen:

Jul 22, 2021 · 2 shares · 2,009 followers View post →
EMDB - EMPIAR @EBI
@EMDB_EMPIAR · X

Check out the paper here! https://t.co/nz6l8Z0Zkf

Jul 22, 2021 · 1 like · 6,122 followers View post →