ReGen Ventures’ Post

🌟Attended an incredibly engaging 3-hour online workshop on "AI and Machine Learning for Drug Discovery" hosted by Decode Life ! 🌟 The workshop offered a comprehensive exploration of molecular dynamics (MD) simulations, guiding us through each vital step and demonstrating how to perform these simulations using the GROMACS software. The hands-on experience in setting up and running MD simulations was particularly valuable, highlighting their potential in optimizing and predicting drug interactions. A highlight was the introduction to Decode's innovative bioinformatics platform, 'Analogue' by Growdea Technologies Pvt Ltd . This cutting-edge tool harnesses AI and machine learning to enhance the precision and efficiency of drug discovery processes. Many thanks to Decode Life for delivering such an in-depth and forward-thinking session. I'm eager to apply these advanced techniques and explore how they can drive innovation in drug discovery. 💡 #AIML #DrugDiscovery #MolecularDynamics #Bioinformatics #GROMACS #Innovation #DecodeLife #LifelongLearning #Research

Five years of lab work, hundreds of experiments, and now sit on a goldmine of untapped data - ready to be transformed. The Nobel Prize winners in Chemistry 2024 showcased the power of AI in molecular science, sparking your curiosity, but you don't know where to begin. At Ingenie Bio, we bridge the gap between raw data and research discoveries so you don't have to navigate AI complexities alone. Our founder, Fabien Plisson, Ph.D., has spent countless hours in the lab synthesizing analogs from natural products, small molecules, and peptides. By mastering statistics, data science, and chemo/bioinformatics, he turned complex datasets into insights, leading to potent kinase inhibitors and GPCR peptide ligands. How much time could you save with our expert services? If you are ready to turn your data into biomolecular discoveries, drop 'START' below or contact us at https://lnkd.in/gpg2ruQB #AI #MachineLearning #ProteinDesign #DrugDiscovery #Biotech

🌟 I’ve successfully completed the digital modules of the One Day International Workshop on AI and Machine Learning for Drug Discovery, hosted by Decode Life. 🧬🔍 A big shout-out to the brilliant speakers, Avinash Mishra and Shreyansh Suyash, for their insightful presentations on Molecular Dynamic Simulations, QSAR (Quantitative Structure-Activity Relationship), and Drug Molecule Generation. 💡 As a Biotech enthusiast, this experience has deepened my understanding of how these cutting-edge technologies in Bioinformatics can revolutionize drug development. 🚀 Looking forward to applying these insights to innovative projects in the field! 💪 #AI #MachineLearning #DrugDiscovery #Bioinformatics #Workshop #LifeSciences

Accelerated #drug #discovery using #AI 🤖🧬💊!! #ImproveThePatientExperience #SaveLives #LessSideEffects 🙌🫶❤️ “The technology uses complex physics-based calculations to generate the data needed to train SandboxAQ’s LQMs based on various inputs including real-world observations and physical properties of various molecular structures. These models can predict molecular behavior and generate molecular structures with desired properties.” “While LLMs work well for generating text and software code, making a difference in the drug discovery, chemicals, and materials industries requires more complex models. “We have to invent new things that aren’t close to the training dataset. And this is where you actually have to ground the AI in the laws of physics,” [Nadia Harhen, general manager for SandboxAQ’a AI simulation division] said.” ““The LLMs and AI that are out there are really only able to predict data that’s very close to its training dataset. You can’t make leapfrog discoveries using LLMs or traditional AI,” she said. “We want to know if it’s not going to work as quickly as possible so that we can optimize those models and keep moving versus finding out seven years down the road. That’s the beauty of [SandboxAQ’s] compute paradigm.””

📊 𝐅𝐨𝐥𝐝𝐢𝐧𝐠 𝐭𝐡𝐞 𝐇𝐮𝐦𝐚𝐧 𝐏𝐫𝐨𝐭𝐞𝐨𝐦𝐞 𝐔𝐬𝐢𝐧𝐠 𝐁𝐢𝐨𝐍𝐞𝐌𝐨 📊   A significant study published in 𝑺𝒄𝒊𝒆𝒏𝒕𝒊𝒇𝒊𝒄 𝑫𝒂𝒕𝒂, a journal from the Nature Portfolio, introduces an extensive dataset of predicted protein structures for 42,042 distinct human proteins. This comprehensive dataset, derived from the UniProt reference proteome UP000005640, aims to enhance the quality and comparability of structural models for machine learning purposes. 🔸 𝐊𝐞𝐲 𝐈𝐧𝐬𝐢𝐠𝐡𝐭𝐬: The dataset combines state-of-the-art modeling tools including AlphaFold 2, OpenFold, and ESMFold, provided within NVIDIA’s BioNeMo platform, as well as homology modeling using Innophore’s CavitomiX platform. This fusion of methods ensures a high level of structural diversity and accuracy. 🔷 𝐃𝐢𝐯𝐞𝐫𝐬𝐞 𝐌𝐨𝐝𝐞𝐥𝐢𝐧𝐠 𝐓𝐞𝐜𝐡𝐧𝐢𝐪𝐮𝐞𝐬: The dataset integrates various AI-guided structure prediction tools and homology modeling to provide atomistic 3D models for nearly all human protein sequences. 🔷 𝐇𝐢𝐠𝐡-𝐐𝐮𝐚𝐥𝐢𝐭𝐲 𝐒𝐭𝐫𝐮𝐜𝐭𝐮𝐫𝐚𝐥 𝐃𝐚𝐭𝐚: Models have been refined to remove low-confidence regions, include potential ligand-binding sites, and undergo energy minimization to ensure high reliability. 🔷 𝐀𝐩𝐩𝐥𝐢𝐜𝐚𝐭𝐢𝐨𝐧𝐬 𝐢𝐧 𝐃𝐫𝐮𝐠 𝐃𝐢𝐬𝐜𝐨𝐯𝐞𝐫𝐲: The dataset is invaluable for applications such as structure-based drug design, prediction of protein function, and interactions. 🔸 𝐈𝐦𝐩𝐚𝐜𝐭 𝐨𝐧 𝐑𝐞𝐬𝐞𝐚𝐫𝐜𝐡: This dataset represents the most comprehensive collection of human protein structures available today. It is expected to facilitate diverse applications, from structural bioinformatics to machine learning, significantly advancing our understanding of human proteins in health and disease. 🔗 𝐑𝐞𝐚𝐝 𝐭𝐡𝐞 𝐟𝐮𝐥𝐥 𝐚𝐫𝐭𝐢𝐜𝐥𝐞: https://lnkd.in/gcTY_-Hs   #Proteomics #Bioinformatics #AIResearch #DrugDiscovery #HumanProteome ♻️ Repost this if you found it insightful. Share with anyone interested in the latest advancements in protein structure prediction! ♻️ 🌐 Swenson Pharmacy Advisors (SPA) Website: https://lnkd.in/gWFbVAjz

🚀 AlphaFold 3: DeepMind's Next Leap in Molecular Structure Prediction🌐 I know there is a lot of expectation with DeepMind's latest innovation—AlphaFold 3 (AF3). In Evolving Therapeutics we are thrilled with this cutting-edge technologies improving healthcare and innovation. 📈 But how it differenciates from previous versions? ⏭ AF3 builds on AlphaFold 2's success with a simpler yet more flexible architecture: - 🔄 Reduced Dependence on MSAs: AF3 relies less on multiple sequence alignments, using a simpler approach to model pairwise relationships between different parts of a molecule. - 🧘♂️ Generative Diffusion Module: AF3 uses a diffusion-based process to predict raw atom coordinates, allowing for a range of possible structures and improving flexibility. - ⏳ Streamlined Structure Prediction: Unlike AlphaFold 2's complex Evoformer module, AF3's Pairformer module offers a more straightforward way to process molecular interactions. ✨ But, why it matters? These differences allow AF3 to handle a wider variety of biological structures with increased accuracy and efficiency, including better understanding protein-protein, protein-RNA, protein-DNA and complexes with small molecules. This opens up new possibilities for biotechnology and medical research. 🧐 Learn More If you're as fascinated by AI and molecular biology as I am, check out these links to learn more about AlphaFold 3 and how it's shaping the future of science: - 📜 https://lnkd.in/dkAthYvW - 🌎 https://lnkd.in/d_QXg6WJ If you are interested in other groundbreaking technologies in the field of biotechnology, look out! 👉 Evolving Therapeutics #ArtificialIntelligence #MolecularBiology #AlphaFold3 #DeepLearning #Biotechnology #EvolvingTherapeutics

🚀 𝐀𝐝𝐯𝐚𝐧𝐜𝐢𝐧𝐠 𝐃𝐫𝐮𝐠-𝐓𝐚𝐫𝐠𝐞𝐭 𝐈𝐧𝐭𝐞𝐫𝐚𝐜𝐭𝐢𝐨𝐧 𝐏𝐫𝐞𝐝𝐢𝐜𝐭𝐢𝐨𝐧: 𝐈𝐧𝐭𝐫𝐨𝐝𝐮𝐜𝐢𝐧𝐠 𝐊𝐨𝐦𝐞𝐭 𝐚𝐧𝐝 𝐋𝐂𝐈𝐝𝐛 𝐃𝐚𝐭𝐚𝐬𝐞𝐭 Drug discovery is entering a new era with the Komet Algorithm and the LCIdb Dataset! Designed to tackle large-scale drug-target interactions (DTIs), these innovations aim to streamline the drug development process and improve predictive accuracy. 𝐊𝐞𝐲 𝐇𝐢𝐠𝐡𝐥𝐢𝐠𝐡𝐭𝐬: 🔬 LCIdb Dataset: A comprehensive DTI dataset with 271,180 molecules and 396,798 positive interactions, offering unparalleled coverage of chemical and biological spaces. This dataset enables robust machine learning training, ensuring high-quality predictions. ⚡ Komet Algorithm: Leveraging the power of the Kronecker Optimized Method, Komet is scalable and efficient. It employs: 1️⃣ GPU-parallel computation for speed. 2️⃣ Feature encoding using the Nyström approximation to reduce memory usage. 3️⃣ Tensor products to capture intricate molecule-protein interactions. 𝐖𝐡𝐲 𝐈𝐭 𝐌𝐚𝐭𝐭𝐞𝐫𝐬: 1️⃣ Improved Predictions: Achieving state-of-the-art performance on both medium-sized and large datasets like DrugBank and LCIdb. 2️⃣ Scalability: Tackles computational challenges head-on, enabling predictions across vast chemical and protein spaces. 3️⃣ Real-World Impact: Demonstrates success in scaffold hopping tasks, critical for identifying novel drug candidates. 𝐀 𝐋𝐞𝐚𝐩 𝐅𝐨𝐫𝐰𝐚𝐫𝐝: With applications ranging from target de-orphanization to off-target identification, Komet and LCIdb mark a major step forward in computational chemogenomics. 𝐄𝐱𝐩𝐥𝐨𝐫𝐞 𝐍𝐨𝐰: 1️⃣ LCIdb and Komet Documentation 👉 https://lnkd.in/dkndinGh 2️⃣ Dataset: Zenodo Repository 👉 https://lnkd.in/dkvziSA5 Let’s shape the future of drug discovery! 🚀 Follow Medvolt for more #DrugDiscovery #AIInHealthcare #MachineLearning #Biotechnology #Innovation #KometAlgorithm #Chemogenomics

Generative biology isn’t the future—it’s happening now. On November 5, Michael Nally, our CEO, will take the stage at the Association of Strategic Alliance Professionals' 2024 BioPharma Conference to reveal how we will use machine learning and computational tools to accelerate drug discovery. In his keynote, “Entering the Era of Generative Biology to Accelerate Drug Discovery,” Mike will break down how strategic partnerships and cutting-edge technology will drive success rates, reduce costs, and deliver better medicines faster than ever. The future of medicine starts here: https://lnkd.in/ePnMAmFM #GenerativeAI | #ProgrammingBiology

A seismic shift is taking place in structural biology! DeepMind's AI software, AlphaFold3, is a game-changer in predicting protein interactions and modifications. Moving beyond mere protein shape prediction, it models molecular interactions between proteins and other compounds, such as DNA, RNA, and small ligands. This breakthrough development can empower scientists with a deeper grasp of life processes, revolutionizing fields like bio-renewables, crop resistance, drug development, and our very understanding of life itself. AlphaFold3's enhanced computation courtesy of its keen machine-learning and extensive training data. Allegedly, it's about 50% more accurate than the current software in predicting protein structures and their interactions. Despite being no longer 'open source', it allows up to 20 jobs per day on the free AlphaFold Server. AlphaFold3's potential area of influence is vast, but do you think the move from 'open-source' will limit its potential for discovery or enhance its effective use? #DeepMind #AIResearch https://lnkd.in/erDuP9aG

🚀 RAPIDDOCK: Revolutionizing Molecular Docking in Drug Discovery! 🧬💊 🔷Exciting advancements are underway in drug discovery, and #RAPIDDOCK is leading the charge! By leveraging advanced deep learning techniques, RAPIDDOCK accelerates the molecular docking process by up to 100x, making it a game-changer for researchers aiming to predict molecular interactions with speed and precision. 🔷Comprehensive Embeddings: RAPIDDOCK captures detailed data from ligands and proteins, using advanced embeddings that represent atomic, charge, and amino acid information, enriched by ESM-2 protein language models. 🔷Transformer Network: A robust 16-layer Transformer applies self-attention to identify crucial molecular interaction points, enhancing the accuracy of docking predictions. 🔷Distance Prediction: RAPIDDOCK predicts precise molecular distances, with masking techniques during training for improved reliability across various structures. 🔷Efficient Ligand Positioning: Based on predicted distances, it accurately reconstructs the ligand’s position relative to the protein, delivering faster, more precise docking results. #DrugDiscovery #AIinBiotech #MolecularDocking #RAPIDDOCK #Innovation #Bioinformatics #tools #LinkedInScience