See the future of R&D with Dotmatics! Our 2024 Mini-Summit, featuring insights from CEO Thomas Swalla, is now available on-demand. Watch now to see how Luma can be a game changer for your R&D workflows and drive AI initiatives at your organization: https://lnkd.in/ecXJq_HW
Dotmatics��� Post
More Relevant Posts
-
@ Artificial Intelligence, Deep Learning, Machine Learning
Welcome to "The Molecule AI Odyssey" In the previous post, we delved into our inaugural module, "𝐓𝐀𝐆𝐌𝐨𝐥." We discovered that 𝐭𝐫𝐚𝐝𝐢𝐭𝐢𝐨𝐧𝐚𝐥 𝐝𝐫𝐮𝐠 𝐝𝐞𝐬𝐢𝐠𝐧 relies heavily on screening libraries to identify viable lead candidates, yet often generates impractical molecules that fall short against established criteria for successful drugs. 𝐓𝐀𝐆𝐌𝐨𝐥 offers a 𝐠𝐫𝐨𝐮𝐧𝐝𝐛𝐫𝐞𝐚𝐤𝐢𝐧𝐠 𝐬𝐨𝐥𝐮𝐭𝐢𝐨𝐧 by employing an 𝐢𝐧𝐧𝐨𝐯𝐚𝐭𝐢𝐯𝐞 𝐨𝐩𝐭𝐢𝐦𝐢𝐳𝐚𝐭𝐢𝐨𝐧 𝐭𝐞𝐜𝐡𝐧𝐢𝐪𝐮𝐞 that not only creates drugs but also enhances multiple properties concurrently. Furthermore, 𝐓𝐀𝐆𝐌𝐨𝐥 achieved recognition at the [ICML] Int'l Conference on Machine Learning 𝐌𝐋4𝐋𝐌𝐒 𝐰𝐨𝐫𝐤𝐬𝐡𝐨𝐩, a prestigious gathering within the deep learning domain. In this module we will introduce you to our second state of the art capability: “𝐌𝐀𝐈-𝐀𝐃𝐌𝐄𝐓: 𝐇𝐢𝐠𝐡𝐥𝐲 𝐚𝐜𝐜𝐮𝐫𝐚𝐭𝐞 𝐀𝐃𝐌𝐄𝐓 𝐩𝐫𝐞𝐝𝐢𝐜𝐭𝐢𝐨𝐧 𝐜𝐨𝐯𝐞𝐫𝐢𝐧𝐠 39 𝐀𝐃𝐌𝐄𝐓 𝐞𝐧𝐝𝐩𝐨𝐢𝐧𝐭𝐬.” 𝐏𝐫𝐞𝐝𝐢𝐜𝐭𝐢𝐨𝐧 𝐨𝐟 𝐭𝐡𝐞 𝐀𝐃𝐌𝐄𝐓 𝐩𝐫𝐨𝐩𝐞𝐫𝐭𝐢𝐞𝐬 of potential drug molecules is vital in the drug-discovery pipeline. Our 𝐡𝐢𝐠𝐡𝐥𝐲 𝐚𝐜𝐜𝐮𝐫𝐚𝐭𝐞 𝐀𝐃𝐌𝐄𝐓 prediction models allow users to bypass the time-consuming and expensive ADMET assays in a wet lab. We cover 19 𝐀𝐃𝐌𝐄 and 20 𝐭𝐨𝐱𝐢𝐜𝐢𝐭𝐲 endpoints. These cover widely-used and well-accepted ADMET properties. We use a 𝐬𝐭𝐚𝐭𝐞-𝐨𝐟-𝐭𝐡𝐞-𝐚𝐫𝐭 graph neural network architecture and over 200 computationally derived molecular features as inputs to the network, over and above the native molecular features. As an illustration of the power of our model, we present an evaluation of the predictive performance of the toxicity prediction of 𝐌𝐀𝐈-𝐀𝐃𝐌𝐄𝐓, which we will refer to as 𝐌𝐀𝐈𝐓𝐨𝐱 here onwards, compared to other freely accessible toxicity prediction tools. 𝘛𝘩𝘦 𝘳𝘦𝘴𝘶𝘭𝘵𝘴 𝘢𝘯𝘥 𝘵𝘦𝘴𝘵 𝘤𝘢𝘴𝘦𝘴 𝘸𝘰𝘶𝘭𝘥 𝘣𝘦 𝘥𝘪𝘴𝘤𝘶𝘴𝘴𝘦𝘥 𝘪𝘯 𝘵𝘩𝘦 𝘯𝘦𝘹𝘵 𝘱𝘰𝘴𝘵. 𝘛𝘪𝘭𝘭 𝘵𝘩𝘦𝘯 𝘴𝘵𝘢𝘺 𝘵𝘶𝘯𝘦𝘥 𝘸𝘪𝘵𝘩 “𝘛𝘩𝘦 𝘔𝘰𝘭𝘦𝘤𝘶𝘭𝘦 𝘈𝘐 𝘖𝘥𝘺𝘴𝘴𝘦𝘺” . #ai #drugdiscovery #machinelearning #pharmacy #ICML #technology
-
-
Associate Director, Data Science and AI Platforms - Applying machine learning to enhance the drug discovery process
Hello, I would like to invite you to Data Science Summit (14.06.2024) where I'm going to talk about "Enhancing Drug Discovery with Machine Learning: Molecular Property Prediction" Full agenda ➡️ https://ml.dssconf.pl/ See you soon! #ArtificialIntelligence #MachineLearning #DataScience #DeepLearning #AI #BigData #DrugDiscovery #PropertyPrediction #ADME #Tox #CADD
-
-
This partnership between Databricks and TetraScience fills the gap that currently exists between siloed scientific data and achieving scientific AI for improved and accelerated outcomes Scientific AI use cases are the atomic unit of value creation where analytics/ML/AI meets engineered scientific data and designed to fulfill scientific needs/outcomes that are not currently possible or even imagined This partnership is centered around scientists, enabling them to derive insights from vast amounts of historical and current data, make faster decisions, and eliminate tedious and error prone tasks
We’re excited to announce our new strategic partnership with TetraScience! This new collaboration will help customers in the life sciences industry unlock data intelligence and harness the value of Scientific AI. Together, we’ll help customers bring more effective and safer therapies to market faster and more cost-efficiently. https://dbricks.co/3K2E0kt
TetraScience
tetrascience.com
-
@ Machine Learning, Artificial Intelligence, Deep Learning, NLP, IoT, Robotics, Computer Vision, GenAI
Welcome to "The Molecule AI Odyssey" In the previous post, we delved into our inaugural module, "𝐓𝐀𝐆𝐌𝐨𝐥." We discovered that 𝐭𝐫𝐚𝐝𝐢𝐭𝐢𝐨𝐧𝐚𝐥 𝐝𝐫𝐮𝐠 𝐝𝐞𝐬𝐢𝐠𝐧 relies heavily on screening libraries to identify viable lead candidates, yet often generates impractical molecules that fall short against established criteria for successful drugs. 𝐓𝐀𝐆𝐌𝐨𝐥 offers a 𝐠𝐫𝐨𝐮𝐧𝐝𝐛𝐫𝐞𝐚𝐤𝐢𝐧𝐠 𝐬𝐨𝐥𝐮𝐭𝐢𝐨𝐧 by employing an 𝐢𝐧𝐧𝐨𝐯𝐚𝐭𝐢𝐯𝐞 𝐨𝐩𝐭𝐢𝐦𝐢𝐳𝐚𝐭𝐢𝐨𝐧 𝐭𝐞𝐜𝐡𝐧𝐢𝐪𝐮𝐞 that not only creates drugs but also enhances multiple properties concurrently. Furthermore, 𝐓𝐀𝐆𝐌𝐨𝐥 achieved recognition at the [ICML] Int'l Conference on Machine Learning 𝐌𝐋4𝐋𝐌𝐒 𝐰𝐨𝐫𝐤𝐬𝐡𝐨𝐩, a prestigious gathering within the deep learning domain. In this module we will introduce you to our second state of the art capability: “𝐌𝐀𝐈-𝐀𝐃𝐌𝐄𝐓: 𝐇𝐢𝐠𝐡𝐥𝐲 𝐚𝐜𝐜𝐮𝐫𝐚𝐭𝐞 𝐀𝐃𝐌𝐄𝐓 𝐩𝐫𝐞𝐝𝐢𝐜𝐭𝐢𝐨𝐧 𝐜𝐨𝐯𝐞𝐫𝐢𝐧𝐠 39 𝐀𝐃𝐌𝐄𝐓 𝐞𝐧𝐝𝐩𝐨𝐢𝐧𝐭𝐬.” 𝐏𝐫𝐞𝐝𝐢𝐜𝐭𝐢𝐨𝐧 𝐨𝐟 𝐭𝐡𝐞 𝐀𝐃𝐌𝐄𝐓 𝐩𝐫𝐨𝐩𝐞𝐫𝐭𝐢𝐞𝐬 of potential drug molecules is vital in the drug-discovery pipeline. Our 𝐡𝐢𝐠𝐡𝐥𝐲 𝐚𝐜𝐜𝐮𝐫𝐚𝐭𝐞 𝐀𝐃𝐌𝐄𝐓 prediction models allow users to bypass the time-consuming and expensive ADMET assays in a wet lab. We cover 19 𝐀𝐃𝐌𝐄 and 20 𝐭𝐨𝐱𝐢𝐜𝐢𝐭𝐲 endpoints. These cover widely-used and well-accepted ADMET properties. We use a 𝐬𝐭𝐚𝐭𝐞-𝐨𝐟-𝐭𝐡𝐞-𝐚𝐫𝐭 graph neural network architecture and over 200 computationally derived molecular features as inputs to the network, over and above the native molecular features. As an illustration of the power of our model, we present an evaluation of the predictive performance of the toxicity prediction of 𝐌𝐀𝐈-𝐀𝐃𝐌𝐄𝐓, which we will refer to as 𝐌𝐀𝐈𝐓𝐨𝐱 here onwards, compared to other freely accessible toxicity prediction tools. 𝘛𝘩𝘦 𝘳𝘦𝘴𝘶𝘭𝘵𝘴 𝘢𝘯𝘥 𝘵𝘦𝘴𝘵 𝘤𝘢𝘴𝘦𝘴 𝘸𝘰𝘶𝘭𝘥 𝘣𝘦 𝘥𝘪𝘴𝘤𝘶𝘴𝘴𝘦𝘥 𝘪𝘯 𝘵𝘩𝘦 𝘯𝘦𝘹𝘵 𝘱𝘰𝘴𝘵. 𝘛𝘪𝘭𝘭 𝘵𝘩𝘦𝘯 𝘴𝘵𝘢𝘺 𝘵𝘶𝘯𝘦𝘥 𝘸𝘪𝘵𝘩 “𝘛𝘩𝘦 𝘔𝘰𝘭𝘦𝘤𝘶𝘭𝘦 𝘈𝘐 𝘖𝘥𝘺𝘴𝘴𝘦𝘺” . #ai #drugdiscovery #machinelearning #pharmacy #ICML #technology
-
-
We believe Scientific {Data & AI} will radically improve and extend human life, by bringing more effective and safer therapies to market faster and less expensively. It is all underpinned by driven people doing the hard work to bring these technologies to bear in the form of solutions that are needed today. With that mission and a great team, it's been amazing to watch TetraScience's journey since founding back in 2015 out of MIT & Harvard where we met them as part of the seed investment round at Underscore VC. They have now brought life that vision as a suite of next-generation lab data management products for scientific use cases, starting with biopharma. Today, they are announcing a market-shaping strategic partnership with Databricks to help life sciences organizations harness that power. As Bavesh Patel, SVP, Go To Market & Industry at Databricks put it: "TetraScience stands out for its unique combination of expertise in science, data, and AI and its resulting ability to deliver Scientific AI outcomes. By aligning that strength with our industry-leading Data Intelligence Platform, we’re establishing the definitive architecture for yielding tangible gains from scientific data across the pharmaceutical value chain.” Well done, Patrick Grady and Siping Wang, for bringing this together and excited to see what your partnership has already brought.
TetraScience + Databricks = The Scientific AI Revolution Today, these two incredible companies announced their strategic partnership dedicated to helping life sciences organizations harness Scientific AI to bring more effective and safer therapies to market faster and less expensively. This is what we love to call a "win-win-win" with customers (and humanity) being the greatest winners of all. Read more:
Accelerating the Scientific AI Revolution: TetraScience and Databricks Join Forces To Transform Scientific Research, Development, Manufacturing, and Quality Control in Life Sciences
tetrascience.com
-
Pave an efficient path from target identification to preclinical development with the power of Logica’s integrated, AI-driven system. We go beyond stand-alone #AI, backed by an expansive experimental power and industry-leading modeling. Visit our website to unlock the potential of #Logica for your #drugdiscovery journey. https://okt.to/BpPlWS #logica #drugdiscovery #AI
AI Drug Discovery | Logica®
https://www.logica.ai
-
We are excited to announce ADMET-AI, a new AI/ML platform for large-scale #ADMET prediction. ADMET-AI is free and open-source with an easy-to-use web interface at https://lnkd.in/gXgiQrhY. See below for more details, and see our bioRxiv paper: https://lnkd.in/gB6DFDRa In #drugdiscovery, predicting the Absorption, Distribution, Metabolism, Excretion, and Toxicity (ADMET) profiles of small molecules is crucial to selecting a small set of compounds with favorable druglike properties for experimental validation. As high-throughput docking and generative AI methods have greatly expanded the number of compounds under consideration in drug discovery pipelines, the speed of ADMET prediction is increasingly critical. In collaboration with Kyle Swanson and James Zou of Stanford University, we developed ADMET-AI, an AI/ML platform for fast and accurate ADMET prediction. ADMET-AI is accurate. It has the highest average rank on the TDC ADMET leaderboard (https://lnkd.in/gPgAV4pR). ADMET-AI is fast. The ADMET-AI web interface is the fastest web-based ADMET predictor, with a 45% speedup compared to the next fastest ADMET web server. ADMET-AI can also be run locally as a Python package, with predictions for 1,000,000 molecules taking just 3.1 hours. Please see these references for more details! Website: https://lnkd.in/gXgiQrhY Code: https://lnkd.in/gK7zYbD3 Paper: https://lnkd.in/gB6DFDRa #drugdevelopment #pharma #ai
-
-
Artificial Intelligence and Computational Sciences Recruitment Consultant UK/EU at Barrington James
"Each stage of the drug discovery process generates large datasets that have to be collected, managed, and analyzed to extract insights that support downstream research and development and guide decision making. In this timely GEN webinar, Dr. Chetanya Pandya will discuss contemporary methods and challenges in managing complex drug discovery data, including novel modalities and applying AI in drug discovery." #drugdiscovery #AI #data #engineering #datasets #machinelearning #development https://lnkd.in/eU_gaJU2
Optimizing Drug Discovery: Managing Complex Data and AI Applications
genengnews.com
-
https://lnkd.in/gKCqkZ4C Now generate novel small molecules that match target product profiles (TPPs) using GenAI just like generating text and images
Merck finds drug discovery DALL-E, becoming early user of small molecule generative AI tool
fiercebiotech.com
-
More high-quality data in the open is needed to enable computational #drugdiscovery. In the last 20 years, SGC has been sharing protein structures and molecular tools using #openscience to accelerate drug discovery. Now, we're embarking on a new phase. Our goal? To produce high-quality, #machinelearning -readable datasets for thousands of human proteins. We're thrilled to collaborate with X-Chem, Inc., utilizing their DNA-Encoded Library (DEL) platform. As Marie-Aude Guié rightly pointed out "If we want the AI community to continue supporting the drug discovery community, we must support the AI community too". https://lnkd.in/g8C32SDR
X-Chem and the SGC are pioneering crowd-sourced AI advancements by making DEL screening data public. | X-Chem
https://www.x-chemrx.com
