TeraLab – Statistics, AI, HPC & Bioinformatics Research Group @ Sapienza’s Post

Federated Learning for Dynamic Resource Allocation in Biomedical Data Analytics Details: This paper proposes a novel federated learning framework for dynamically allocating computational resources in biomedical data analytics. Current approaches to analyzing large-scale biomedical datasets, such as genomic sequencing and medical imaging, often suffer from resource bottlenecks and inefficient data utilization. Our framework, termed "Adaptive Federated Resource Orchestrator" (AFRO), combines federated learning techniques with dynamic resource allocation algorithms to optimize computational resource utilization across distributed computing nodes. AFRO utilizes a centralized orchestrator that leverages reinforcement learning to dynamically adjust the resource allocation strategy based on real-time performance metrics, ensuring efficient training of federated models across heterogeneous data sources. Originality: AFRO distinguishes itself by integrating reinforcement learning for adaptive resource allocation within a federated learning context – a combination not wide https://lnkd.in/geVcCGhn

🚀 Generative AI isn’t just coming to medicine — it’s already here. A new 2025 review in Annual Reviews shows how multimodal models (text + imaging + data) are transforming clinical work: 🧠 drafting medical notes 🩻 reconstructing scans 🔬 generating synthetic data for research The challenge? Trust, validation, and ethical deployment — not technology. AI won’t replace doctors. But it will redefine what great medicine looks like. 👉 Generative Artificial Intelligence in Medicine – Annual Reviews (2025) https://lnkd.in/dDSK_9pn #GenerativeAI #MedicalAI #HealthcareInnovation #DigitalHealth

🔥 Generative AI is no longer a buzzword — it’s transforming medicine right now. From drafting reports to reconstructing scans and generating synthetic data — the shift from analysis to creation has begun. At Labwise AI sp. z.o.o, we’re seeing the same evolution: integrating lab, genetic, and wearable data to predict health outcomes before symptoms appear. The question is no longer if AI will change healthcare — but how responsibly we’ll let it. 💡 🔗 Generative Artificial Intelligence in Medicine – Annual Reviews (2025) https://lnkd.in/dQy24jhF #GenerativeAI #MedicalAI #DigitalHealth #LabWiseAI #HealthcareInnovation

🚀 AI and Genomics: Shaping the Next Generation of Biomedical Innovators https://www.chip.org/ Boston Children’s Hospital’s Computational Health Informatics Program (CHIP) continues to set the gold standard in AI-driven genomics research and training. Their Postdoctoral Fellowship in AI and Genomics, led by Dr. Kenneth D. Mandl and colleagues at Harvard Medical School, embodies a powerful vision: Integrating clinical, molecular, genomic, and environmental data using AI to drive precision medicine and healthcare transformation This program trains scientists to bridge data science, medicine, and genomics, producing leaders who go on to faculty positions and groundbreaking research careers. It’s a model of what the future of biomedicine looks like — multimodal, data-integrative, and AI-enabled. https://www.chip.org/ 🌱 At BioDigitEra, we share this same commitment — but our focus is on making this frontier accessible to life scientists around the world. Our Genomic Data Science and AI for Life Scientists program is designed to: 🧬 Empower wet-lab biologists to transition into computational biology 💻 Teach real-world data analysis using Python, R, and cloud-based genomics tools ☁️ Integrate AI, ML, and big data into genomic interpretation and research 🌍 Build a global network of trained bioinformaticians driving discovery in Africa, Asia, and beyond Because the promise of precision medicine depends not only on technology — but on people who can translate data into understanding, and understanding into care. The next decade of science belongs to those who can connect genomics, AI, and human health. Let’s prepare together for that future. #Genomics #AI #Bioinformatics #DataScience #BiomedicalInformatics #PrecisionMedicine #BioDigitEra #Harvard #BostonChildrensHospital #ComputationalBiology #LifeScience #Training #Innovation

Scientists are working on 'wetware' to run the compute powering AI - tiny human 'brains' grown in a lab and attached to electrodes. Slightly uncomfortable reading perhaps, but not something new. One of the podcasts we recorded for the VistaMilk research centre - "The Rise of the Bionanomachines" – talked about programming microbes to deliver messages and their use in organic computing. (https://lnkd.in/edwHZKWC) #organiccomputing #biotech #AI https://lnkd.in/dwQdtAHH

Recent developments in computational biology have enabled the design of intrinsically disordered proteins (IDPs), which are challenging to predict using conventional AI tools. A new machine learning approach leverages physics-based simulations and automatic differentiation algorithms to design IDPs with tailored properties. This method allows for precise optimization of protein sequences, providing insights into protein behavior and potential applications in disease research and therapeutics. By integrating molecular dynamics with advanced computational techniques, this approach represents a significant step forward in the rational design of complex biomolecules.

Glad to share that our paper "A Hybrid #Graph and #LLM Approach for Measuring #Scientific #Novelty via Knowledge Recombination and Propagation" has been published by Expert Systems with Applications (Impact Factor = 7.5). In this paper, we: 1️⃣ Propose a hybrid graph and LLM framework for measuring scientific novelty via fine-grained knowledge recombination and propagation model. 2️⃣ Combine content-based and reference-based approaches for comprehensive novelty quantification. 3️⃣ Conduct extensive experiments on top AI and BME conferences and verify the effectiveness and stability of our proposed model. 4️⃣ Reveal knowledge and novelty patterns distinguishing awarded and non-award papers. We will release the data and code soon at: https://lnkd.in/g_6TKRYG UNT College of Information

✨Exploring the Intersection of AI, Bioinformatics and Education✨ A year ago, we shared a vision for AI in Bioinformatics and Education and it’s more relevant than ever!!! Our publication, “No-Boundary Thinking for Artificial Intelligence in Bioinformatics and Education” was a collaborative effort with Dr(s) Nisha Pillai and Prajay Patel, where we delved into how AI is reshaping bioinformatics and education. We discussed: *Innovative Teaching Approaches: Integrating AI tools to enhance learning experiences. *Curriculum Evolution: Adapting educational frameworks to include AI-driven methodologies. *Future Prospects: Envisioning the role of AI in the next generation of bioinformatics professionals. This work reflects our commitment to bridging the gap between advanced technology, computational sciences and education, fostering a new wave of scientific inquiry. The best ideas happen when we remove the boundaries between disciplines. 🔬📚 🔗 Read the full article here: https://lnkd.in/gVwjhgqw #AIinEducation #Bioinformatics ##Datascience #MachineLearning #EducationalInnovation #Stem #STEMEducation

🧬 Workshop on Artificial Intelligence in Biology | Organized by Biotecnika 🤖 Thrilled to have attended an enriching workshop that explored how Artificial Intelligence is revolutionizing Biology and Life Sciences! 🌿💡 The 17-day workshop covered: ✅ Fundamentals of AI, Machine Learning & Deep Learning ✅ Applications of AI in genomics, drug discovery, and molecular biology ✅ How algorithms analyze and predict complex biological patterns ✅ Insights into the future of precision medicine and data-driven research From Machine Learning to Deep Learning, the sessions showcased how AI can transform biological research — enabling smarter data analysis, accelerating discoveries, and shaping the future of biotechnology. This certification course beautifully bridges Biology and Technology, inspiring me to explore how computational intelligence can drive innovation in the life sciences. #AIinBiology #ArtificialIntelligence #Biotecnika #Biotechnology #MachineLearning #DeepLearning #Bioinformatics #LifeSciences #Innovation #Research #Learning

Just recorded with Ron Weiss at MIT - one of the pioneers of synthetic biology who's been at this since 1996, back when he was helping set up a wet lab in MIT's CS department. The conversation really clarified something for me about why synthetic biology is finally working. Early on, Ron and others tried to build digital logic circuits in cells - AND gates, NOT gates, the whole thing. Computer scientists thought it made perfect sense. Biologists thought they'd come from outer space. ("Why would you want to make bacteria blink?") But digital logic doesn't scale in biology. You hit a wall around 4-input gates. Cells are already running near full capacity. Digital circuits need massive protein overproduction to maintain clear on/off states. The cells resist. They mutate your carefully designed DNA. They push back. The breakthrough wasn't abandoning the computational approach. It was realizing biology already computes - just not digitally. It uses analog signals, graded responses, intermediate values. More like neural networks than logic gates. We can now start to build actual neural networks inside cells. Perceptrons that do weighted calculations using RNA and proteins. These networks can be designed by biocompilers in silico and compiled into DNA sequences. Ron's work on self-amplifying RNA with Jacob Becraft and Tasuku Kitada led to Strand Therapeutics. One RNA molecule becomes 200,000 copies, each carrying both therapeutic payload and computational logic to detect cancer cells. They started first-in-human trials just over a year ago. Some melanoma patients - who'd exhausted all other options - have had complete remissions. The REACT project with Ken Shepard pushes even further: implantable devices that use optogenetics to control engineered cells, with your iPhone as the interface. One day, you might be able to literally dial up or down therapeutic protein production. We spent decades trying to make therapeutics that compute digitally, in binary, using yes and no, presence or absence. Turns out biology was already computing in a far more sophisticated way. We just had to learn its language. Ron started programming mainframes with punch cards at age 7. Now he's teaching living cells to think. This is the century of biology. And it's just the beginning. Full episode out now on frameshifts.bio.