📣 Excited to share that “MoltGraph: A Longitudinal Temporal Graph Dataset of Moltbook for Coordinated-Agent Detection” was accepted to the 1st ACM Conference on AI and Agentic Systems! 📄 Paper: arxiv.org/pdf/2603.00646 #AgenticAI #GraphMining #BotDetection #CAIS

PredPPI-GReMLIN: Prediction of Protein-Protein Interactions through Mining of Conserved Bipartite Graphs 1. A novel graph-based framework, predPPI-GReMLIN, has been introduced to predict protein-protein interactions (PPIs) by integrating atomic-level physicochemical descriptors and conserved substructure mining. This method models PPI interfaces as bipartite graphs, capturing both spatial and structural details that are often overlooked by traditional sequence-based approaches. 2. The core innovation of predPPI-GReMLIN lies in its ability to detect conserved interaction patterns within protein interfaces using a graph-search strategy. By matching these patterns against a large corpus of known protein complexes, the method can predict novel interaction partners and ligands with high accuracy. 3. Evaluations on diverse datasets, including CAMP, Yeast, and TAGPPI, demonstrate superior performance compared to state-of-the-art methods. predPPI-GReMLIN achieves over 97% precision, recall, accuracy, and F1-scores in binary classification tasks, highlighting its robust predictive capabilities. 4. A significant advancement is the incorporation of solvent-accessible surface area (SASA) features, which enhances the accuracy of multi-class interaction type classification to 57.74%. This integration provides deeper insights into the physicochemical context of PPIs. 5. A case study on the SARS-CoV-2 spike-ACE2 complex further validates the method. Docking simulations using predicted ligands reproduce native-like binding energetics, with statistically significant results compared to random ligands (p=5.1×10^-9). This highlights the potential for drug discovery applications. 6. The dataset and source code for predPPI-GReMLIN are publicly available, enabling researchers to explore and build upon this innovative framework. This accessibility is crucial for advancing the field of structural bioinformatics and PPI prediction. 💻Code: github.com/morufwork/predPPI… 📜Paper: biorxiv.org/content/10.64898… #ProteinInteraction #GraphMining #Bioinformatics #MachineLearning #StructuralBiology #PPIPrediction

"Graph neural networks in TensorFlow" from the TensorFlow Blog 👉blog.tensorflow.org/2024/02/… #DataScience #GraphMining #MachineLearning

📢 Our recent survey, "A Survey on Hypergraph Mining: Patterns, Tools, and Generators," with Geon, Fanchen, and @tinaeliassi is now on ArXiv. Link: arxiv.org/abs/2401.08878 Your feedback is highly appreciated. Please send me an email. #NetworkScience #GraphMining

💻Grafik Sinir Ağları (GNNs), yapay zeka alanında kullanılan ve grafik veri yapıları üzerinde işlem yapabilen derin öğrenme modelleridir. Bu ağlar, bir grafikte bulunan düğümlerin ve kenarların özelliklerini dikkate  alarak;  komşuluk ilişkilerini inceleyip bilgi paylaşımını kolaylaştırmakta ve özelliklerin güncellenmesini sağlamaktadır. Grafik Sinir Ağları (GNNs) hakkında bir #bioinfonotes paylaşılmıştır. İngilizce bilimsel içerikler içeren @bioinfocodes hesabımızı takip edebilirsiniz. #GNN  #graphs #nodes #embeddings #convolution #isomorphism #aggregation #classification #prediction #learning #representation #nodeembeddings #messagepassing #graphattentionnetworks #graphmining #structure #network #AcademicTwitter #AcademicTwitter #bilimlekalalım

A very interesting paper by Lutz Oettershagen, @nmkriege, @chrsmrrs and Petra Mutzel on static representations of temporal graphs doi.org/10.1137/1.9781611976… . #GraphML #GraphMining #NetworkScience #MachineLearning

🥳Congratulations are in order for @Stanford Professor & former KGC keynote speaker Jure Leskovec! 👏@Jure won the #SIGKDD Innovation Award for his contributions to applied #MachineLearning, #GraphMining & #NetworkScience! 📺Watch his KGC Keynote ⬇️: youtu.be/T4afZsyHmUc

🎉 Excited and honored to receive the #SIGKDD Innovation Award for my contributions to #GraphMining, #NetworkScience, and applied #MachineLearning! Looking forward to advancing these fields even further! 🚀 I am grateful for all the amazing students and collaborators. Thank you to all my mentors and everyone who supported me on this journey! 🙌 #KDD2023 @kdd_news

We have an open research internship position on the design of deep GNNs! Please help spread the word! #GNN #GraphMachineLearning #MachineLearning #GraphMining

ProbGraph, a project led by Maciej Besta, won the Best Paper Award at #SC22. ProbGraph is a graph representation that enables simple and fast approximate parallel #graphmining with strong theoretical guarantees on work, depth, and result accuracy. #HPC spcl.inf.ethz.ch/Publication…

Best paper award @kdd_news goes to Alibaba team for Federated GNN. Congrats. :) Interesting to see #GraphML or #GraphMining does not seem to be a primary subject area in the submission form. :D #KDD2022

⚠️ New paper! ⚠️ "Discovering Polarization Niches via Dense Subgraphs with Attractors and Repulsers" accepted at @VLDBconf #vldb #graphmining #AcademicTwitter

We have multiple Applied #MachineLearning Scientist openings in our team. Come join us in San Diego and Seattle to work on exciting #GraphMining, #TimeSeries, #NLP and #ComputerVision challenges impacting millions of people across the world: amazon.jobs/en/jobs/1867937/…

Check out @Tanmoy_Chak course on Social Network Analysis. The course starts today, 25 Jul 2022. Course webpage: lnkd.in/dMbQwSTf Course teaser: lnkd.in/d5dDBEJe #SocialNetworkAnalysis #OSNs #GraphMining

The #deadline for research papers has been extended to June 23. Time to wrap up those awesome papers and submit! Great ideas, findings and results on social networks analysis are welcome. Please retweet! @ASONAM_News #ASONAM2022 #socialmedia #datamining #graphmining #BigData

Learning the concept of #cluster and #spanner helps you to better understand your graph. Here is me looking for them in Obsidian.😉 #obsmd #graphmining Click to learn more about them.😍

📢 Open Assistant Professor position in #AI at @centralesupelec @UnivParisSaclay @Inria_Saclay! Please contact me if you are working in graph mining and learning, and you're interested to apply. #NetworkScience #GraphMining #GNN #AcademicTwitter 🔗centralesupelec.fr/sites/def…

Had a blast virtually giving a talk at the @Vanderbilt_CS ML Lunch seminar on #tensors for #multiaspect #graphmining Thanks so much @TylersNetwork and @SKolouri and everyone who attended for hosting and great discussions!

This evening thinking about this brilliant GraphINVENT Package by @rociomer3 Graph neural networks for molecular design (please check out latest addition of reinforcement learning framework) - github.com/MolecularAI/Graph… #ai #ml #datascience #healthcare #drugdiscovery #graphmining

Follow us on LinkedIn to see all of our long-form posts on #AI, #graphdatabases, #graphanalytics and #graphmining: bit.ly/3D9fb1e