Scores by kind of claim
Each kind of claim is scored on its own, out of 100: the share that held up, with the range it likely falls in. Fewer resolved claims give a wider range.
-
Factual claims
Statements about the present or the past, checked against the record: did they hold up?
81 76 87 -
Predictions
Statements about events outside the person's control, checked once they fell due: did they come true?
Insufficient record: 2 predictions resolved, 5 needed for a score.
-
Promises
Commitments to the person's or organisation's own future action, with something to deliver: was it delivered?
65 45 83
All types combined: 80 (likely 75 to 85, high confidence), from 84 claims: 78 factual claims, 6 promises. It mixes kinds of claim in different proportions for different people, so compare people kind by kind.
Demis Hassabis's claims
37 claims
“Today, in partnership with EMBL-EBI, we’re incredibly proud to be launching the AlphaFold Protein Structure Database, which offers the most complete and accurate picture of the human proteome to date, more than doubling humanity’s accumulated knowledge of high-accuracy human protein structures.”
Demis Hassabis, DeepMind CEO, on DeepMind blog 22 July 2021 Putting the power of AlphaFold into the world's hands (Google DeepMind)
DeepMind and EMBL-EBI launched the AlphaFold Protein Structure Database, the most complete and accurate picture of the human proteome to date, more than doubling known high-accuracy human protein structures.
Resolved true The record shows it happened or was so.
Said for an organisation they lead or work for; it counts on both records.
What would count as true or false
Judge as of 2021-07-22. True: DeepMind and EMBL-EBI launched the AlphaFold Protein Structure Database around that date, covering about the whole human proteome, and the high-confidence human structures it added more than doubled the share of residues or proteins that previously had experimental structures (e.g. Tunyasuvunakool et al., Nature 2021). Mostly true: the launch happened as stated, but the 'more than doubling' figure holds only on a narrower measure, or comes close to doubling without exceeding it. Mixed: the launch happened, but the 'most complete and accurate' or 'more than doubling' parts are substantially unsupported. Mostly false: the database launched, but the coverage claims are clearly overstated. False: no such joint launch took place.
Evidence
- Highly accurate protein structure prediction for the human proteome - Nature Nature (Nature Publishing Group) · 22 July 2021 Own record
“Of the human proteome, 35.7% of total residues fall within the highest accuracy band (corresponding to 38.6% of residues for which a prediction was produced) (Fig. 1c). This is double the number of residues covered by an experimental structure.”
- Highly accurate protein structure prediction for the human proteome - Nature Nature (Nature Publishing Group) · 22 July 2021 Own record
“After decades of effort, 17% of the total residues in human protein sequences are covered by an experimentally determined structure1. Here we markedly expand the structural coverage of the proteome by applying the state-of-the-art machine learning method, AlphaFold2, at a scale that covers almost the entire human proteome (98.5% of human proteins).”
- DeepMind and EMBL release the most complete database of predicted 3D structures of human proteins | EMBL EMBL (European Molecular Biology Laboratory) · 22 July 2021
“The database dramatically expands the accumulated knowledge of protein structures, more than doubling the number of high-accuracy human protein structures available to researchers.”
- Accessible 3D protein models to accelerate scientific discovery EMBL-EBI · 22 July 2021
“In collaboration with EMBL-EBI, it has made the AlphaFold protein predictions, source code and methodology freely and, crucially, openly available to the global scientific community through the AlphaFold database.”
Reasoning
EMBL and EMBL-EBI both reported on 2021-07-22 that DeepMind, working with EMBL-EBI, had released the AlphaFold Protein Structure Database (E3, E4). EMBL's announcement says the database more than doubled the number of high-accuracy human protein structures (E3). The Nature paper says the predictions cover 98.5% of human proteins (E2). It also puts 35.7% of human residues in the highest-accuracy band, against 17% previously covered by experimental structures (E1, E2). That is about 2.1 times the earlier coverage, so it meets the residue-based doubling test in the criteria.
Cite this claim
Demis Hassabis, “DeepMind and EMBL-EBI launched the AlphaFold Protein Structure Database, the most complete and accurate picture of the human proteome to date, more than doubling known high-accuracy human protein structures.” (said 22 July 2021). Verdict: resolved true. Reliability API, claim c_4fe80884f883, method 0.1.0+c876030c. https://epistacklabs.com/claims/c_4fe80884f883
Claims without a verdict
Is a verdict here wrong, or do you speak for Demis Hassabis? Reply on the claim's own page, with the form at its foot (how replies work). Verdicts are made automatically; how verdicts and scores are made.