Michael Smith’s Post

"Pharma companies have always honed their competitive edge through creativity and ingenuity. AI won’t change that." Interesting insights about AI and the future of life sciences manufacturing. The single most important part of developing a comprehensive AI strategy is having the best dataset possible - garbage in, garbage out. Rockwell's DataMosiax platform helps our customers enable controlled access to relevant and contextualized data in a scalable fashion. The basis for implementing any AI must start with excellent data, and we can help you leverage your data into actionable knowledge! #AI #machinelearning

Over the past several years, the problem of data cleaning became visible to me in a pretty visceral way. Data Scientists and PhDs in biostatistics devoting significant energy to cleaning data instead of advancing scientific efforts. I joined Cornerstone AI because I met an exceptionally skilled and passionate technical and scientific team equally frustrated with the status quo of data cleaning. We perform Data Quality Assessments, Data Standardization & Error Detection, and Data Pipeline Optimization faster, with higher quality, and likely more cost efficiency than current methods. #lifesciences #realworlddata #datacleaning #ai

#AI is already transforming drug discovery research. But to truly unlock its potential, it all starts with having an AI-ready data strategy. Discover the essentials to ensure your data is optimized.

Sharing Bob's post below with my network. This excellent eBook included a "cut down" version of the Museum Article Bob McDowall and I wrote....

When I joined SAS, I was motivated by the immense potential of SAS Viya to accelerate breakthroughs across the life sciences value chain. Today, AI and machine learning are revolutionizing how we solve complex business challenges, and it's clear that faster, more collaborative insights are critical for advancing drug development, clinical trials, and real-world evidence analysis.   Recent evaluations by The Futurum Group confirmed that SAS Viya delivers 4x greater productivity in managing the full AI/ML lifecycle—far outpacing competing platforms. This kind of productivity boost not only enables life sciences organizations to derive deeper insights faster but also helps optimize resource allocation and reduce costs across the board.   With the ability to scale insights globally and tailor our AI capabilities to diverse data types, SAS Viya is setting a new standard for innovation in the pharmaceutical and biotech industries. I’m proud to be part of a team that’s empowering data and AI teams to focus on breakthroughs, rather than being bogged down by complexity. #LifeSciences #AI #SASViya #Innovation #Pharma #HealthcareAI #DataAnalytics

Laboratory of the Future eBook A new eBook on Lab of the Future has just been published by Technology Networks. It contains interesting articles including: 1. A shortened version of #PaulSmith and my Museum of Analytical Antiquities that we have both posted recently 2. Top Ten Compliance Tips by me. This looks at what are the main compliance features you should be considering as laboratories move to the future. 3. How AI and machine learning are streamlining data handling and advancing research. I should point out that that AI is Artificial Intelligence not Insemination. People wanting the latter topic should consider alternative sites on the Internet. 4. Sustainable lab practices that reduce environmental impact without sacrificing research quality The eBook is available from: https://lnkd.in/e9AdeRUG

A fantastic start to the week: OpenFold today released two new AI tools, SoloSeq and OpenFold-Multimer. OpenFold is an incredible model for the potential of public-private partnerships, and for the responsible use of AI tools to further science across industries. These powerful technologies stand to enhance our ability to design high-quality and effective crop protection products, and additionally they exemplify Bayer's commitment to open innovation and industry collaboration. The integration of AI tools like these into our workflows augments the efforts of #TeamBayer scientists, enabling us to move faster and more efficiently.   These tools tie seamlessly into Bayer's CropKey initiative, propelling us further in our mission to revolutionize the way crop protection products are developed and designed. By leveraging the open-source nature of these tools, we can fine-tune models with our proprietary data, driving new scientific breakthroughs. This is a tremendous step forward in our journey to reshape agriculture through BioAI. Let's continue to harness the power of AI to unlock the full potential of nature in a sustainable way. #AI #BioAI #AgTech #CropProtection 

Data is the “business of lifeblood” as specialist firms like TetraScience work to accelerate scientific AI by designing and industrializing AI-native scientific datasets.Food for thoughts, Shaping Expert AI path #AI #Biopharma #Lifescience #acceleration

𝐓𝐡𝐨𝐮𝐠𝐡𝐭𝐬 𝐟𝐨𝐫 𝐓𝐡𝐮𝐫𝐬𝐝𝐚𝐲 𝘉𝘶𝘪𝘭𝘥𝘪𝘯𝘨 𝘍𝘰𝘶𝘯𝘥𝘢𝘵𝘪𝘰𝘯𝘢𝘭 𝘋𝘢𝘵𝘢 If you want to build foundational data, build the foundation to enable your scientists. Your scientists will ask the questions to enable AI. As AI is pushing forward, the concept of foundational data in increasing in popularity. I think it is great, but at the same time, it is reactive in nature to enable AI. Way before AI, the scientists that I worked with were trying to ask questions of the data... "If I increase (specific component) in my feed, will it reduce my end of run lactate spike?" We had put the data into several types of database and neural network technologies to attempt to ask questions of the data. Granted we were ahead of our time because we put the business of the business before the newest thing, but we had gained an understanding of "what goes where". Santha Ramakrishnan PhD mentions that we should "Plan for access and integration to data so models are not just built on PowerPoint Think of how data will be used multimodally when you plan for management of select data domains". We must take a step back from our own "world" and understand that our data ties into a bigger picture, from research, to development, to the vivarium, to the clinic, to commercial... through the cloning, through the cycles and functional groups, to pharmacokinetic and pharmacodynamic. When you are building your foundation... Build it for your scientists... Build it for the patients, at the end of the day, we are the patients too. Then build it for AI.

[𝗔𝗜 𝗛𝗶𝘀𝘁𝗼𝗿𝗶𝗲𝘀] 📚✨ 𝗘𝘅𝗽𝗲𝗿𝘁 𝗦𝘆𝘀𝘁𝗲𝗺 An expert system is an AI system that embodies specific domain knowledge to solve problems within that domain, such as medical diagnosis, financial services, customer support, and more. An expert system is a type of knowledge-based AI system, among others like Case-Based Reasoning (CBR) Systems, Frame-Based Systems, Semantic Networks, Ontology-Based Systems, Fuzzy Logic Systems, and Bayesian Networks. It has five principal components: the knowledge base, inference engine, user interface, explanation facility, and knowledge acquisition module. The core component is the knowledge base. The most common and straightforward way to represent knowledge is by rules—a set of if-then expressions, such as "IF the soil moisture level is low THEN water the plant." The IF part is called the 𝘢𝘯𝘵𝘦𝘤𝘦𝘥𝘦𝘯𝘵, and the THEN part is the 𝘤𝘰𝘯𝘴𝘦𝘲𝘶𝘦𝘯𝘵. We can combine several antecedents to derive one consequent, such as "IF the soil moisture level is low AND the weather forecast is rainy THEN do not water the plant." One of the earliest expert systems is 𝗗𝗘𝗡𝗗𝗥𝗔𝗟 (short for Dendritic Algorithm), developed at Stanford University in 1965 by a group of researchers led by Edward Feigenbaum, who is often regarded as the "father of expert systems." The aim of the project was to study hypothesis formation and discovery in science. A particular use case was chosen: helping chemists identify unknown molecules using mass spectra and chemistry knowledge. Rules have many advantages. They are: - Fast - Direct - Explicit - Easy to interpret and explain the decision process The last point is a crucial one for an AI system aiming to provide recommendations, as it allows for gaining user confidence and adoption. Rule-based systems are different from decision trees. In rule-based systems, rules are independent, making it easy to add, remove, or change rules, though it becomes difficult to maintain as it scales. Decision trees can be considered as chained rules: each consequent is also an antecedent for the next rule. Following DENDRAL, among famous projects were 𝗠𝗬𝗖𝗜𝗡 in the 1970s, designed for diagnosing bacterial infections and recommending antibiotics, and 𝗥𝟭/𝗫𝗖𝗢𝗡 in the 1980s, used for configuring VAX computer systems. Expert systems made intangible knowledge a tangible asset. They proved useful, could outperform human experts, and could generate significant commercial value. Thus, this period marked the 𝘀𝗲𝗰𝗼𝗻𝗱 𝗔𝗜 𝗯𝗼𝗼𝗺. #machinelearning #artificialintelligence #datascience #ml #ai -------- In this series, I explore the history of AI to help you better understand its evolution. You'll learn about the challenges faced at each stage and the solutions developed. Many of these challenges persist today, and some of the methods and strategies are still in use. As progress builds on past advancements, I hope this series also inspires your next solution.