Eric and Wendy Schmidt Center’s Post

Extracting information about cell states from images — also called image-based profiling — can shed light on disease mechanisms and gene functions. Machine learning has not yet been fully integrated into these approaches. Srinivas Niranj Chandrasekaran, Shantanu Singh, Anne Carpenter, and colleagues from the JUMP Cell Painting Consortium created a dataset, called CPJUMP1, of three million annotated cell images and used it to compare genetic and chemical perturbations targeting the same gene. They found that identifying perturbations that produce the same observable change in cells was challenging, and lay a framework for developing methods that identify cell features in images. Read more in Nature Methods. #BroadInstitute #Science #ScienceNews #Research #ScientificResearch #CellPainting

Identifying similarities between genetic and chemical perturbations is often challenging. The group of Anne Carpenter from Broad Institute of MIT and Harvard have built this dataset to profile and compare the morphological profiles of cells with genetic and chemical perturbations (160 genes and 303 compounds) based on the microscopy images. Great work, congratulations to all the authors! #BroadInstitute #Science #Research #drugscreening #screening #morphology #cellprofiler #imageanalysis

Generative AI arrives in the gene editing world of CRISPR Now, new AI technology is generating blueprints for microscopic biological mechanisms that can edit your DNA, pointing to a future when scientists can battle illness and diseases with even greater precision and speed than they can today. Read more at: https://lnkd.in/dMtmQCrq

Researchers at MIT have developed a method to study gene regulatory networks using observational data, bypassing the need for costly or unethical interventional experiments. The approach leverages machine learning (ML), specifically a causal disentanglement algorithm, to identify and aggregate related genes into functional modules, revealing their cause-and-effect relationships. ML plays a critical role by efficiently processing complex genetic data and uncovering hidden patterns that traditional methods might miss. This technique promises to advance drug development and precision medicine, offering a more scalable and cost-effective way to understand gene interactions and develop targeted treatments for diseases. https://lnkd.in/ekhyxXau

<<first multimodal deep learning model of its kind, EPBDxDNABERT-2, capable of ascertaining the precise relationship between transcription factors, proteins that regulate gene activities, leveraging an aspect of DNA called DNA breathing, in which the double-helix structure opens and closes spontaneously. The model has the potential to aid in the design of drugs used to treat diseases that originate in gene activity.>>

🧬 Groundbreaking News: AI Revolutionizes CRISPR Gene Editing Thrilled to share the latest breakthrough in genetic engineering: OpenCRISPR-1, the first AI-designed gene editor, has demonstrated remarkable precision with a 95% reduction in off-target effects compared to traditional methods. This innovation, developed by Profluent Bio, represents a paradigm shift in how we approach genetic engineering. The system combines large language models trained on 5.1 million Cas9-like proteins to create entirely synthetic, yet highly effective gene-editing tools. What makes this particularly exciting: 🎯 Comparable efficiency to SpCas9 with superior precision 🧪 Fully synthetic molecules designed from scratch by AI 🔓 Open-source availability for ethical research and development 🔬 Validated performance in human cell experiments The implications for precision medicine are profound, especially as whole genome sequencing costs continue to decrease. We're witnessing the convergence of AI and biotechnology in ways previously confined to science fiction. Looking forward to seeing how this technology advances therapeutic applications and genetic research. What are your thoughts on AI-designed biological tools? #Biotechnology #AI #CRISPR #Innovation #GeneticEngineering #Science

🔬 Revolution in Gene Editing: 🧬 As highlighted in a recent New York Times article, Profluent, a pioneering startup from Berkeley, is leveraging generative AI to transform CRISPR technology—already famous for its ability to edit DNA with high precision. This collaboration between AI and CRISPR promises to enhance genetic engineering, paving the way for more targeted and efficient therapies. Profluent's approach, which combines the power of AI learning with the precision of CRISPR, opens up new possibilities for curing genetic diseases and advancing personalized medicine. 🌐 Discover more about Profluent’s groundbreaking integration of AI with CRISPR and their vision for a healthier world in the full article from the New York Times: https://lnkd.in/gW6Gs4wQ #CRISPR #GeneticEngineering #AI #NewYorkTimes #MedicalInnovation

"Advances in gene sequencing technology and computing power have significantly increased the availability of bioinformatic data and processing capabilities. This convergence provides an ideal opportunity for artificial intelligence (AI) to develop methods to control cellular behavior." #bioinformatics #gene #genomics #sequencing #machinelearning #AI

🔥 BREAKING: Profluent Bio has cracked the code by combining AI with gene editing. Meet OpenCRISPR-1, the game-changer in genetic engineering! 🔥 🔹 AI-Driven Design: Utilizing Large Language Models (LLMs) trained on massive datasets of CRISPR and Cas9 proteins, OpenCRISPR-1 was engineered from a database of 5.1 million Cas9-like proteins and 238,917 Cas9 proteins from the CRISPR-Cas Atlas. 🔹 Unmatched Precision: This AI-created gene editor features a Cas9-like protein and guide RNA, boasting comparable or improved efficiency over traditional CRISPR systems like SpCas9. The results? Higher specificity and minimal DNA damage, thanks to a reduction in off-target effects. 🔹 Public Access for All: In a groundbreaking move, Profluent Bio has made OpenCRISPR-1 publicly available. This democratizes gene editing, supercharging research and commercial applications across the globe. Why should you care? Because this could be the key to unlocking new treatments for genetic diseases. Imagine a safer, faster, and more personalized approach to medicine. Custom gene editors tailored to individual DNA aren't sci-fi anymore—they’re on the horizon because of innovations like OpenCRISPR-1. 💥 The Future of Medicine is Now. Don’t get left behind. What genetic disorder would you target first with OpenCRISPR-1? Let’s get a discussion going! 👇 #AI #Genetics #Innovation

I just published "Fourier Neural Operators in Gene Transcription" In this blog, I explore why Fourier Neural Operators (FNOs) provide a more effective and scalable solution than transformers for modeling the complex, multi-scale processes of gene transcription, from capturing high-frequency regulatory features to long-range genomic interactions #Genomics #ArtificialIntelligence #Math https://lnkd.in/gSq3aafi