Bo Wang
@BoWang87
SVP and Head of Biomedical AI @Xaira_Thera; Associate Prof @UofT; Chief AI Officer @UHN; former PHD, CS @Stanford; opinions my own. #AI #healthcare #biology
Thrilled to share that we placed 1st in the 2025 MEDIQA Clinical NLP Challenge — marking our third consecutive win (2023, 2024, 2025)! 🏆🎉 This year’s task went beyond clinical note generation — it focused on extracting structured medical orders from noisy, free-form…
🔥Just wrapped up an incredible three days in Montreal at the AACR (@AACR ) Special Conference on Artificial Intelligence and Machine Learning — the first AACR event fully dedicated to exploring how AI is reshaping cancer research and care. I was deeply honored to serve as one of…
It’s incredible that this IMO gold medal–level model isn’t using any specialized mathematical proof tools like Lean, but instead relies entirely on pure textual reasoning. This strongly reinforces the foundation of our BioReason model (arxiv.org/abs/2505.23579 ), which applies…
1/N I’m excited to share that our latest @OpenAI experimental reasoning LLM has achieved a longstanding grand challenge in AI: gold medal-level performance on the world’s most prestigious math competition—the International Math Olympiad (IMO).
🚀 Excited to share our latest protocol in @NatureProtocols : "scGPT: End-to-end Protocol for Fine-tuned Retinal Cell Type Annotation"! Single-cell RNA-seq is transforming biology, but annotating large-scale, high-resolution data is challenging. Our team introduces a powerful…
Our team is at #ICML2025 ! Please visit our poster if you are interested to know more about our work MEDRAX!
MedRAX is at #ICML2025 and we have candy! Come see us at W305 @JunMa_11 @BoWang87 @UHN #AIHealth #Radiology
Thanks for highlighting some of my research @uhnfoundation !
Inspired by tools like ChatGPT, @UHN researchers are creating powerful AI models that can read and understand DNA, RNA and proteins — all at the same time — to better predict how cells behave and how diseases develop ➜ bit.ly/44uZIZB @UHN_Research @BoWang87
🚀 how good is Grok-4 on biomedical applications ? Yesterday, @xai benchmarked Grok-4 on our Chest Agent benchmark—and it crushed it. With 72.8% accuracy, Grok-4 outperformed our MedRAX (63.1%) (the previous STOA) and all other models in chest X-ray interpretation. 🔗 MedRAX:…
Just had the pleasure of being interviewed by @Nature on the rise of AI in digital pathology. With growing workloads and global shortages of pathologists, the field is turning to AI not as a luxury—but as a necessity. Full article: nature.com/articles/d4158… In the article, I…
scRNAseq cell type annotation is notoriously messy. Despite so many algorithms, most researchers still rely on manual annotations using marker genes In a new preprint accepted at ICML GenAI Bio Workshop, we ask if reasoning LLMs (DeepSeek-R1) can help with cell type annotation🧵
🎉 Huge milestone: MedSAM has officially surpassed 2,000 citations! 🏆 MedSAM was the first foundation model for medical imaging! When we open-sourced MedSAM, our goal was to empower the entire medical imaging community—and seeing it become a global standard is beyond…
I usually don't talk about preprints, but I will make an exception here: Introducing #MedSAM (Segment Anything in Medical Images arxiv.org/abs/2304.12306) for universal medical image segmentation! 📷 1st attempt adapting SAM to medical domains 📷 Large-scale supervised training…
Really liked this meme from @lucascamillomd 😅
Do deep generative models in single-cell omics really work for perturbation prediction? Some benchmark studies say yes: 🔗 arxiv.org/pdf/2408.10609 🔗biorxiv.org/content/10.110… Others say no: 🔗 arxiv.org/abs/2410.13956 🔗 BMC Genomics: bmcgenomics.biomedcentral.com/articles/10.11… To move beyond this…
Do deep generative models in single-cell omics really work for perturbation prediction? Some benchmark studies say yes: 🔗 arxiv.org/pdf/2408.10609 🔗biorxiv.org/content/10.110… Others say no: 🔗 arxiv.org/abs/2410.13956 🔗 BMC Genomics: bmcgenomics.biomedcentral.com/articles/10.11… To move beyond this…
🚀 What do genomic Transformers actually learn about biology? •What knowledge do they hold at random init, after pre‑training, and following fine‑tuning? •We dove deep into every attention head to find out. 📄 Preprint live now “Interpreting Attention Mechanisms in Genomic…
Exciting day for the virtual cell revolution! 🚀 @arcinstitute has just launched the Virtual Cell Challenge, calling on leading institutions worldwide to push the boundaries in virtual cell modeling (virtualcellchallenge.org). Huge congrats to Arc Team (@genophoria @pdhsu…
Register today for the Virtual Cell Challenge and use AI to solve one of biology’s most complex problems. Announced in @CellCellPress, the competition is hosted by Arc Institute and sponsored by @nvidia, @10xGenomics, and @UltimaGenomics.
What a mind-blowing week for AI in biology! 🚀 Xaira just dropped the largest genome-wide perturbation dataset ever last week 🧬 Arc unveiled STATE yesterday, the most powerful foundation model for single-cell perturbation prediction 🧠 DeepMind followed up with AlphaGenome,…
More in-depth technical tweetorial from the amazing @_annhuang , the first author of our X-Atlas/Orion paper! x.com/_annhuang/stat…
🚀 Xaira Therapeutics has just dropped a game-changer for AI-driven biology. Today, we unveiled X-Atlas/Orion, the largest publicly available genome-wide Perturb-seq dataset to date—spanning 8.4 million single cells with perturbations across all ~20,000 human protein-coding…