PubHawk
Nat Commun March 2023 PMID 36859488

Predicting locations of cryptic pockets from single protein structures using the PocketMiner graph neural network.

Artur Meller, Michael Ward, Jonathan Borowsky, Meghana Kshirsagar, Jeffrey M Lotthammer, Felipe Oviedo, Juan Lavista Ferres, Gregory R Bowman

Full text PDF Download social card

Mentions

19
Peak: 10 in Mar 2023
across 3 sources since Mar 2023

Extended reach

18,351
potential audience

Total likes

21

Video views

30

Mentions per month

Mar 2023 – Jun 2023

Where it's discussed

  • X / Twitter10 posts
  • News8 articles
  • YouTube1 videos

AI synopsis

What people are saying

Discussions consistently praise the publication's innovative approach to identifying cryptic pockets with high accuracy, with words like 'exciting,' 'game-changing,' and 'significant' indicating strong positive sentiment and perceived impact. References to the use of machine learning and the potential for expanding druggable targets highlight both technical interest and practical implications.

AgreementModerate agreement

Most posts express positive acknowledgment and support for the significance of the publication, highlighting its novelty and utility.

InterestHigh interest

Discussions show high curiosity and enthusiasm, referencing the potential for drug discovery and technological advancements.

EngagementModerate engagement

Comments include mentions of specific methods like graph neural networks, indication deep analysis and appreciation.

ImpactHigh impact

Posts emphasize the high potential influence on fields like drug development, computational biology, and AI research.

Abstract in brief

  • PocketMiner is a graph neural network designed to quickly and accurately predict cryptic pockets in proteins, which are hidden sites that can be targeted for drug discovery but are hard to detect.
  • It was tested on a new dataset of confirmed cryptic pockets and shown to identify these pockets with high accuracy (ROCAUC of 0.87) and much faster than previous methods.
  • When applied to the entire human proteome, PocketMiner suggests that many proteins previously thought to lack druggable pockets may actually contain hidden cryptic pockets, significantly expanding the potential targets for drug development.
Read the full abstract

Top posts

Most liked on X and Bluesky
CureCoin
@CureCoin_Team · X

#Foldingathome community acknowledged in latest paper: Predicting locations of cryptic pockets from single protein structures using the PocketMiner graph #neuralnetworks https://t.co/OElyT2ky0u https://t.co/MGBAmT6OAq

Mar 9, 2023 · 8 likes · 3 shares · 7,975 followers View post →
Arnau Comajuncosa-Creus
@acomajuncosa · X

Predicting locations of cryptic pockets from single protein structures using the PocketMiner graph neural network https://t.co/6I7slMoGnv

Mar 21, 2023 · 7 likes · 111 followers View post →
UPenn Dept of Biochem & Biophys
@BB_UPenn · X

@GuptaPenn @PLOSPathogens The Bowman Lab recently published a paper in Nature Communications. Read it here: https://t.co/GHN9RaAx6k https://t.co/EJh33lqoJu

Mar 7, 2023 · 3 likes · 1 share · 1,035 followers View post →
Paul Lopez
@lopezunwired · X

Predicting locations of cryptic pockets from single protein structures using the PocketMiner graph neural network https://t.co/Rqdu2LoalT #MachineLearning #NatureJournal #AI https://t.co/YkTxE642VG

Mar 1, 2023 · 1 like · 2,483 followers View post →
Reluctant Quant
@DrMattCrowson · X

RT Predicting locations of cryptic pockets from single protein structures using the PocketMiner graph neural network https://t.co/R4pKKbYqUY https://t.co/5CmxVqxfvF

Mar 1, 2023 · 1 like · 1,261 followers View post →
Ordo Fraterna Fibonacci
@OrdoFibonacci · X

Predicting locations of cryptic pockets from single protein structures using the PocketMiner graph neural network | Nature Communications https://t.co/AxarBySA1Z

Mar 22, 2023 · 2,568 followers View post →

All mentions

Ordo Fraterna Fibonacci
@OrdoFibonacci · X

Predicting locations of cryptic pockets from single protein structures using the PocketMiner graph neural network | Nature Communications https://t.co/AxarBySA1Z

Mar 22, 2023 · 2,568 followers View post →
Arnau Comajuncosa-Creus
@acomajuncosa · X

Predicting locations of cryptic pockets from single protein structures using the PocketMiner graph neural network https://t.co/6I7slMoGnv

Mar 21, 2023 · 7 likes · 111 followers View post →
tropicalwi
@tropicalwi · X

RT @CureCoin_Team: #Foldingathome community acknowledged in latest paper: Predicting locations of cryptic pockets from single protein str…

Mar 9, 2023 · 3 shares · 506 followers View post →
Ivan Tuma
@IvanTumaJr · X

RT @CureCoin_Team: #Foldingathome community acknowledged in latest paper: Predicting locations of cryptic pockets from single protein str…

Mar 9, 2023 · 3 shares · 2,017 followers View post →
Christophe Verschuere
@TheRealRebel74 · X

RT @CureCoin_Team: #Foldingathome community acknowledged in latest paper: Predicting locations of cryptic pockets from single protein str…

Mar 9, 2023 · 3 shares · 249 followers View post →
CureCoin
@CureCoin_Team · X

#Foldingathome community acknowledged in latest paper: Predicting locations of cryptic pockets from single protein structures using the PocketMiner graph #neuralnetworks https://t.co/OElyT2ky0u https://t.co/MGBAmT6OAq

Mar 9, 2023 · 8 likes · 3 shares · 7,975 followers View post →
NotedSource
@NotedSource · X

RT @BB_UPenn: @GuptaPenn @PLOSPathogens The Bowman Lab recently published a paper in Nature Communications. Read it here: https://t.co/GHN9…

Mar 7, 2023 · 1 share · 88 followers View post →
UPenn Dept of Biochem & Biophys
@BB_UPenn · X

@GuptaPenn @PLOSPathogens The Bowman Lab recently published a paper in Nature Communications. Read it here: https://t.co/GHN9RaAx6k https://t.co/EJh33lqoJu

Mar 7, 2023 · 3 likes · 1 share · 1,035 followers View post →
Paul Lopez
@lopezunwired · X

Predicting locations of cryptic pockets from single protein structures using the PocketMiner graph neural network https://t.co/Rqdu2LoalT #MachineLearning #NatureJournal #AI https://t.co/YkTxE642VG

Mar 1, 2023 · 1 like · 2,483 followers View post →
Reluctant Quant
@DrMattCrowson · X

RT Predicting locations of cryptic pockets from single protein structures using the PocketMiner graph neural network https://t.co/R4pKKbYqUY https://t.co/5CmxVqxfvF

Mar 1, 2023 · 1 like · 1,261 followers View post →