Great to see @eqtlcatalogue.bsky.social data strongly represented in variant interpretation benchmarks in the AlphaMissense preprint (expression QTLs, splicing QTLs and polyadenylation QTLs). Clean, uniformly processed data resources matter! @elixirestonia.bsky.social
Excited to launch our AlphaGenome API goo.gle/3ZPUeFX along with the preprint goo.gle/45AkUyc describing and evaluating our latest DNA sequence model powering the API. Looking forward to seeing how scientists use it! @googledeepmind
June 26, 2025 at 10:51 AM
Great to see @eqtlcatalogue.bsky.social data strongly represented in variant interpretation benchmarks in the AlphaMissense preprint (expression QTLs, splicing QTLs and polyadenylation QTLs). Clean, uniformly processed data resources matter! @elixirestonia.bsky.social
Ranking of Variant Effect Predictor (VEP) in-silico tools from a recently published review, based on Relative performance with rate on MAVE data)
1. CPT-1 (Broad Institute)
2. AlphaMissense (Google DeepMind and cols.)
3. ESCOTT (Sorbonne/INSERM France)
1. CPT-1 (Broad Institute)
2. AlphaMissense (Google DeepMind and cols.)
3. ESCOTT (Sorbonne/INSERM France)
April 16, 2025 at 9:41 AM
Ranking of Variant Effect Predictor (VEP) in-silico tools from a recently published review, based on Relative performance with rate on MAVE data)
1. CPT-1 (Broad Institute)
2. AlphaMissense (Google DeepMind and cols.)
3. ESCOTT (Sorbonne/INSERM France)
1. CPT-1 (Broad Institute)
2. AlphaMissense (Google DeepMind and cols.)
3. ESCOTT (Sorbonne/INSERM France)
A nice paper from Martin Morgan and colleagues in @bioinfoadv.bsky.social | AlphaMissenseR: an integrated framework for investigating missense mutations in human protein-coding genes | #Bioinformatics #AlphaMissense #Alphafold #RStats #Bioconductor 🧬 🖥️ 🧪 🔓
⬇️
academic.oup.com/bioinformati...
⬇️
academic.oup.com/bioinformati...
AlphaMissenseR: an integrated framework for investigating missense mutations in human protein-coding genes
AbstractSummary. AlphaMissense is an AI model from Google DeepMind that predicts the pathogenicity of every possible missense mutation in the human proteom
academic.oup.com
May 1, 2025 at 12:43 PM
A nice paper from Martin Morgan and colleagues in @bioinfoadv.bsky.social | AlphaMissenseR: an integrated framework for investigating missense mutations in human protein-coding genes | #Bioinformatics #AlphaMissense #Alphafold #RStats #Bioconductor 🧬 🖥️ 🧪 🔓
⬇️
academic.oup.com/bioinformati...
⬇️
academic.oup.com/bioinformati...
NEW: Scientists have built an AI genetics analysis model to boost the diagnosis and treatment of rare diseases that are estimated to affect hundreds of millions of people worlwide.
It has already been used to help a muscular atrophy sufferer in Senegal.
www.ft.com/content/bc49...
It has already been used to help a muscular atrophy sufferer in Senegal.
www.ft.com/content/bc49...
New AI model enhances diagnosis of rare diseases
PopEVE system outperformed rivals such as Google DeepMind’s AlphaMissense
www.ft.com
November 24, 2025 at 10:38 AM
NEW: Scientists have built an AI genetics analysis model to boost the diagnosis and treatment of rare diseases that are estimated to affect hundreds of millions of people worlwide.
It has already been used to help a muscular atrophy sufferer in Senegal.
www.ft.com/content/bc49...
It has already been used to help a muscular atrophy sufferer in Senegal.
www.ft.com/content/bc49...
AlphaMissense, building on AlphaFold, based on unsupervised learning AI, predicts impact of all 71 million human genome variants for disease-causing potential, across entire human proteome; open-resource 🧪
science.org/doi/10.1126/science.adg7492
science.org/doi/10.1126/science.adg7492
September 19, 2023 at 3:16 PM
AlphaMissense, building on AlphaFold, based on unsupervised learning AI, predicts impact of all 71 million human genome variants for disease-causing potential, across entire human proteome; open-resource 🧪
science.org/doi/10.1126/science.adg7492
science.org/doi/10.1126/science.adg7492
New preprint! Assessment of AlphaMissense against dozens of other methods. Overall AlphaMissense has most power to detect association of rare coding variants with common clinical phenotypes. But varies a lot between genes - for some, other methods are much better. www.medrxiv.org/content/10.1...
🧪🧬🖥️
🧪🧬🖥️
Assessment of ability of AlphaMissense to identify variants affecting susceptibility to common disea...
medRxiv - The Preprint Server for Health Sciences
www.medrxiv.org
November 27, 2023 at 11:30 AM
New preprint! Assessment of AlphaMissense against dozens of other methods. Overall AlphaMissense has most power to detect association of rare coding variants with common clinical phenotypes. But varies a lot between genes - for some, other methods are much better. www.medrxiv.org/content/10.1...
🧪🧬🖥️
🧪🧬🖥️
PopEVE covered by @financialtimes.com today! www.ft.com/content/bc49...
New AI model enhances diagnosis of rare diseases
PopEVE system outperformed rivals such as Google DeepMind’s AlphaMissense
www.ft.com
November 24, 2025 at 10:23 AM
PopEVE covered by @financialtimes.com today! www.ft.com/content/bc49...
#alphafold AlphaMissense data integrated into Ensembl, UniProt and AlphaFold DB !
alphafold.ebi.ac.uk
alphafold.ebi.ac.uk
May 22, 2024 at 7:03 AM
#alphafold AlphaMissense data integrated into Ensembl, UniProt and AlphaFold DB !
alphafold.ebi.ac.uk
alphafold.ebi.ac.uk
AlphaMissense looks exciting, but I wish every published paper was reproducible and extensible.
"we are not releasing the trained model weights, the code is not meant to be used to reproduce the AlphaMissense scores"
www.science.org/doi/10.1126/...
github.com/deepmind/alp...
"we are not releasing the trained model weights, the code is not meant to be used to reproduce the AlphaMissense scores"
www.science.org/doi/10.1126/...
github.com/deepmind/alp...
Science | AAAS
www.science.org
September 21, 2023 at 8:19 PM
AlphaMissense looks exciting, but I wish every published paper was reproducible and extensible.
"we are not releasing the trained model weights, the code is not meant to be used to reproduce the AlphaMissense scores"
www.science.org/doi/10.1126/...
github.com/deepmind/alp...
"we are not releasing the trained model weights, the code is not meant to be used to reproduce the AlphaMissense scores"
www.science.org/doi/10.1126/...
github.com/deepmind/alp...
“Lightway access to AlphaMissense data that demonstrates a balanced performance of this missense mutation predictor” 🧪🧶🧬
Another evaluation of CFTR AlphaMissense predictions
www.biorxiv.org/content/10.1...
Another evaluation of CFTR AlphaMissense predictions
www.biorxiv.org/content/10.1...
November 3, 2023 at 6:03 AM
“Lightway access to AlphaMissense data that demonstrates a balanced performance of this missense mutation predictor” 🧪🧶🧬
Another evaluation of CFTR AlphaMissense predictions
www.biorxiv.org/content/10.1...
Another evaluation of CFTR AlphaMissense predictions
www.biorxiv.org/content/10.1...
Can AI decode cancer’s hidden genetic flaws?
AlphaMissense accurately flags pathogenic DNA repair mutations in genes like BRCA1/2 and ATM.
Yet its predictions fall short for POLE, proving AI still needs clinical validation before guiding cancer treatment.
pmc.ncbi.nlm.nih.gov/articles/PMC...
AlphaMissense accurately flags pathogenic DNA repair mutations in genes like BRCA1/2 and ATM.
Yet its predictions fall short for POLE, proving AI still needs clinical validation before guiding cancer treatment.
pmc.ncbi.nlm.nih.gov/articles/PMC...
September 28, 2026 at 8:20 PM
Can AI decode cancer’s hidden genetic flaws?
AlphaMissense accurately flags pathogenic DNA repair mutations in genes like BRCA1/2 and ATM.
Yet its predictions fall short for POLE, proving AI still needs clinical validation before guiding cancer treatment.
pmc.ncbi.nlm.nih.gov/articles/PMC...
AlphaMissense accurately flags pathogenic DNA repair mutations in genes like BRCA1/2 and ATM.
Yet its predictions fall short for POLE, proving AI still needs clinical validation before guiding cancer treatment.
pmc.ncbi.nlm.nih.gov/articles/PMC...
A new preprint Spotlight from @genomicspc.bsky.social is live!
"AlphaGenome: a Swiss-army knife for exploring non-coding DNA"
by Judit García-González & Krzysztof Gogolewski
"Deciphering the three billion letters of the human genome has long been biology’s ultimate puzzle..."
shorturl.at/qiCIk
"AlphaGenome: a Swiss-army knife for exploring non-coding DNA"
by Judit García-González & Krzysztof Gogolewski
"Deciphering the three billion letters of the human genome has long been biology’s ultimate puzzle..."
shorturl.at/qiCIk
December 4, 2025 at 2:34 PM
A new preprint Spotlight from @genomicspc.bsky.social is live!
"AlphaGenome: a Swiss-army knife for exploring non-coding DNA"
by Judit García-González & Krzysztof Gogolewski
"Deciphering the three billion letters of the human genome has long been biology’s ultimate puzzle..."
shorturl.at/qiCIk
"AlphaGenome: a Swiss-army knife for exploring non-coding DNA"
by Judit García-González & Krzysztof Gogolewski
"Deciphering the three billion letters of the human genome has long been biology’s ultimate puzzle..."
shorturl.at/qiCIk
未知のヒト遺伝子変異が病気を引き起こす可能性を予測するAIモデル「popEVE」が登場、Google DeepMindのAlphaMissenseよりも優秀
https://gigazine.net/news/20251127-human-disease-genetics-ai-model-popeve/
https://gigazine.net/news/20251127-human-disease-genetics-ai-model-popeve/
未知のヒト遺伝子変異が病気を引き起こす可能性を予測するAIモデル「popEVE」が登場、Google DeepMindのAlphaMissenseよりも優秀
ハーバード大学医学部の研究者が、計算生物学におけるGoogle DeepMindの優位性に挑戦するべく、希少遺伝病をより正確に診断するために設計された新しいAIモデルの「popEVE」を発表しました。
gigazine.net
November 26, 2025 at 11:00 PM
未知のヒト遺伝子変異が病気を引き起こす可能性を予測するAIモデル「popEVE」が登場、Google DeepMindのAlphaMissenseよりも優秀
https://gigazine.net/news/20251127-human-disease-genetics-ai-model-popeve/
https://gigazine.net/news/20251127-human-disease-genetics-ai-model-popeve/
New @genomebrowser.bsky.social tracks highlight the power of generative AI for biology:
🧬 AlphaMissense predicts what amino acid variations will create problems in protein folding
🧬 VarChat conducts lit searches for variants
More are in the works! genomics.wordpress.ucsc.edu/news/2025/02...
🧬 AlphaMissense predicts what amino acid variations will create problems in protein folding
🧬 VarChat conducts lit searches for variants
More are in the works! genomics.wordpress.ucsc.edu/news/2025/02...
February 26, 2025 at 9:59 PM
New @genomebrowser.bsky.social tracks highlight the power of generative AI for biology:
🧬 AlphaMissense predicts what amino acid variations will create problems in protein folding
🧬 VarChat conducts lit searches for variants
More are in the works! genomics.wordpress.ucsc.edu/news/2025/02...
🧬 AlphaMissense predicts what amino acid variations will create problems in protein folding
🧬 VarChat conducts lit searches for variants
More are in the works! genomics.wordpress.ucsc.edu/news/2025/02...
Researchers detail popEVE, an AI model to predict the disease-causing potential of unknown human genetic mutations, and says it beats Google's AlphaMissense (Michael Peel/Financial Times)
Main Link | Techmeme Permalink
Main Link | Techmeme Permalink
November 24, 2025 at 11:25 AM
Researchers detail popEVE, an AI model to predict the disease-causing potential of unknown human genetic mutations, and says it beats Google's AlphaMissense (Michael Peel/Financial Times)
Main Link | Techmeme Permalink
Main Link | Techmeme Permalink
AlphaMissense pathogenicity scores predict response to immunotherapy and enhances the predictive capability of tumor mutation burden https://www.biorxiv.org/content/10.1101/2025.09.28.679078v1
October 1, 2025 at 3:47 PM
AlphaMissense pathogenicity scores predict response to immunotherapy and enhances the predictive capability of tumor mutation burden https://www.biorxiv.org/content/10.1101/2025.09.28.679078v1
Have your own screen?
cLIP (clustered Local Interaction Profiler) clusters partners by the residues they contact: interaction fingerprints, contact frequency with pLDDT and AlphaMissense, binding modes on the structure.
You can check out several examples.
flyark.github.io/LIVIA/clip.html
cLIP (clustered Local Interaction Profiler) clusters partners by the residues they contact: interaction fingerprints, contact frequency with pLDDT and AlphaMissense, binding modes on the structure.
You can check out several examples.
flyark.github.io/LIVIA/clip.html
September 25, 2026 at 2:44 PM
Have your own screen?
cLIP (clustered Local Interaction Profiler) clusters partners by the residues they contact: interaction fingerprints, contact frequency with pLDDT and AlphaMissense, binding modes on the structure.
You can check out several examples.
flyark.github.io/LIVIA/clip.html
cLIP (clustered Local Interaction Profiler) clusters partners by the residues they contact: interaction fingerprints, contact frequency with pLDDT and AlphaMissense, binding modes on the structure.
You can check out several examples.
flyark.github.io/LIVIA/clip.html
AlphaMissense - AlphaFold tuned for variant effect prediction
www.science.org/doi/10.1126/...
Accurate proteome-wide missense variant effect prediction with AlphaMissense
www.science.org/doi/10.1126/...
Accurate proteome-wide missense variant effect prediction with AlphaMissense
Science | AAAS
www.science.org
September 19, 2023 at 2:37 PM
AlphaMissense - AlphaFold tuned for variant effect prediction
www.science.org/doi/10.1126/...
Accurate proteome-wide missense variant effect prediction with AlphaMissense
www.science.org/doi/10.1126/...
Accurate proteome-wide missense variant effect prediction with AlphaMissense
New review out! 🧬🔬 We explore how protein structure is improving variant effect prediction. Advances in structural modelling, including AlphaFold, are powering methods like AlphaMissense, PrimateAI-3D, and CPT-1, leading to better accuracy and interpretability. www.sciencedirect.com/science/arti...
Leveraging protein structural information to improve variant effect prediction
Despite massive sequencing efforts, understanding the difference between human pathogenic and benign variants remains a challenge. Computational varia…
www.sciencedirect.com
February 25, 2025 at 9:50 PM
New review out! 🧬🔬 We explore how protein structure is improving variant effect prediction. Advances in structural modelling, including AlphaFold, are powering methods like AlphaMissense, PrimateAI-3D, and CPT-1, leading to better accuracy and interpretability. www.sciencedirect.com/science/arti...
I know you’re not suggesting this, but I don’t think authors are under any obligation to compare to AlphaMissense because one cannot run this with a user-defined input or test on independent data
November 8, 2023 at 8:12 PM
I know you’re not suggesting this, but I don’t think authors are under any obligation to compare to AlphaMissense because one cannot run this with a user-defined input or test on independent data
AlphaMissense promises powerful AI for pathogenic variant identification. Is it clinically useful? Smart validation approach here against known signatures ascopubs.org/doi/10.1200/... Found with researchbriefing.com
ascopubs.org
June 30, 2025 at 8:01 AM
AlphaMissense promises powerful AI for pathogenic variant identification. Is it clinically useful? Smart validation approach here against known signatures ascopubs.org/doi/10.1200/... Found with researchbriefing.com
GPCRvp, a GPCR-specific variant impact predictor, decodes GLP1R missense variants using molecular dynamics + AI. It sharply reduces false positives vs AlphaMissense and improves prediction of GLP 1 drug response. #GLP1R #GPCR #Diabetes #Obesity link.springer.com/article/10.1... 🔓
The GPCRVP score reliably predicts the impact of GLP1R human variants on receptor function - Diabetologia
Aims/hypothesis The glucagon-like peptide-1 receptor (GLP1R) is a key regulator of glucose homeostasis and body weight, and a major therapeutic target for type 2 diabetes and obesity. Individual disea...
link.springer.com
September 25, 2026 at 12:34 PM
GPCRvp, a GPCR-specific variant impact predictor, decodes GLP1R missense variants using molecular dynamics + AI. It sharply reduces false positives vs AlphaMissense and improves prediction of GLP 1 drug response. #GLP1R #GPCR #Diabetes #Obesity link.springer.com/article/10.1... 🔓
Deciphering gain-of-function from loss-of-function variants with AlphaMissense: A case study with the mechanosensitive PIEZO1 ion channel protein https://www.biorxiv.org/content/10.1101/2025.07.03.662957v1
July 6, 2025 at 6:47 AM
Deciphering gain-of-function from loss-of-function variants with AlphaMissense: A case study with the mechanosensitive PIEZO1 ion channel protein https://www.biorxiv.org/content/10.1101/2025.07.03.662957v1
Yep - another is protein structure modelling with #AlphaFold. I use this daily in my job as a clincial geneticist - it's massively useful (along with AlphaMissense). Lumping it together with recreational slop seems... unfair.
November 30, 2025 at 8:03 AM
Yep - another is protein structure modelling with #AlphaFold. I use this daily in my job as a clincial geneticist - it's massively useful (along with AlphaMissense). Lumping it together with recreational slop seems... unfair.
Deciphering gain-of-function from loss-of-function variants with AlphaMissense: A case study with the mechanosensitive PIEZO1 ion channel protein www.biorxiv.org/content/10.1101/2025.07.03.662957v1 #cryoem
July 6, 2025 at 11:46 AM
Deciphering gain-of-function from loss-of-function variants with AlphaMissense: A case study with the mechanosensitive PIEZO1 ion channel protein www.biorxiv.org/content/10.1101/2025.07.03.662957v1 #cryoem