We are excited to share that the NVIDIA healthcare team have officially returned as co-organizers in Part 2 of the Stanford RNA 3D Folding Competition!
\nAs you may have seen, research collaborators @chrismunley @youhanlee from NVIDIA BioNeMo team and KGMoN member @theoviel already posted RNAPro last week. This is the RNA structure prediction model developed by Nvidia and Kaggle hosts in collaboration with the leading Kaggler teams in the previous Part 1 of the competition, as described in our collaborative preprint.
\nCheck out the templates and code in Theo's post to get access. RNAPro, wrapped around the template-based modeling pipeline developed by @jaejohn, is also the basis of publicly shared notebooks being developed by @jaejohn @nikitakuznetsof @yanecoder and others, which are currently highly competitive on the leaderboards. Go Kaggle!
\nWe wanted to say a huge thank you to @youhanlee, @chrismunley and @theoveil (NVIDIA) and @przemekporebski (HHMI) for all the work that was done on the Data and Validation side to get this competition launched.
\nAs the competition contnues, you may see all of us co-organizers providing some further interesting baseline models and answering questions on forums. We are very excited to have Nvidia's continuing scientific collaboration in tackling this grand challenge in biology.
", "messages": [ { "id": 3397685, "postDate": "2026-01-27T19:23:48.770Z", "content": "We are excited to share that the NVIDIA healthcare team have officially returned as co-organizers in Part 2 of the Stanford RNA 3D Folding Competition!
\nAs you may have seen, research collaborators @chrismunley @youhanlee from NVIDIA BioNeMo team and KGMoN member @theoviel already posted RNAPro last week. This is the RNA structure prediction model developed by Nvidia and Kaggle hosts in collaboration with the leading Kaggler teams in the previous Part 1 of the competition, as described in our collaborative preprint.
\nCheck out the templates and code in Theo's post to get access. RNAPro, wrapped around the template-based modeling pipeline developed by @jaejohn, is also the basis of publicly shared notebooks being developed by @jaejohn @nikitakuznetsof @yanecoder and others, which are currently highly competitive on the leaderboards. Go Kaggle!
\nWe wanted to say a huge thank you to @youhanlee, @chrismunley and @theoveil (NVIDIA) and @przemekporebski (HHMI) for all the work that was done on the Data and Validation side to get this competition launched.
\nAs the competition contnues, you may see all of us co-organizers providing some further interesting baseline models and answering questions on forums. We are very excited to have Nvidia's continuing scientific collaboration in tackling this grand challenge in biology.
", "rawMarkdown": "We are excited to share that the NVIDIA healthcare team have officially returned as co-organizers in Part 2 of the Stanford RNA 3D Folding Competition! \n\nAs you may have seen, research collaborators @chrismunley @youhanlee from NVIDIA BioNeMo team and KGMoN member @theoviel already posted RNAPro last week. This is the RNA structure prediction model developed by Nvidia and Kaggle hosts in collaboration with the leading Kaggler teams in the previous Part 1 of the competition, as described in our [collaborative preprint](https://www.biorxiv.org/content/10.64898/2025.12.30.696949v1). \n\nCheck out the templates and code in [Theo's post](https://www.kaggle.com/competitions/stanford-rna-3d-folding-2/discussion/668412) to get access. RNAPro, wrapped around the template-based modeling pipeline developed by @jaejohn, is also the basis of publicly shared notebooks being developed by @jaejohn @nikitakuznetsof @yanecoder and others, which are currently highly competitive on the leaderboards. Go Kaggle!\n\nWe wanted to say a huge thank you to @youhanlee, @chrismunley and @theoveil (NVIDIA) and @przemekporebski (HHMI) for all the work that was done on the Data and Validation side to get this competition launched. \n\nAs the competition contnues, you may see all of us co-organizers providing some further interesting baseline models and answering questions on forums. We are very excited to have Nvidia's continuing scientific collaboration in tackling this grand challenge in biology.", "votes": 15 }, { "id": 3398843, "postDate": "2026-01-29T19:21:29.457Z", "content": "Not sure this was a good thing. I for one stopped working this competition - fine tuning a massive model not my cup of tea when the model designed exactly for the task.
\nGuess kaggle will be able to measure if a good thing based on level of participation.
", "rawMarkdown": "Not sure this was a good thing. I for one stopped working this competition - fine tuning a massive model not my cup of tea when the model designed exactly for the task.\n\nGuess kaggle will be able to measure if a good thing based on level of participation.", "votes": 2 }, { "id": 3414446, "postDate": "2026-02-26T20:01:55.253Z", "content": "Hi everyone,\nI’m looking to join a team.\nBackground in Pharmacovigilance, AI in Healthcare, and Drug Development.\nExperience with ML applications in clinical safety, bioinformatics, and regulatory frameworks (FDA/EMA).\nInterested in collaborating with strong deep learning / geometric modeling experts.\nHappy to contribute domain expertise, modeling strategy, and validation.
", "rawMarkdown": "Hi everyone,\nI’m looking to join a team.\nBackground in Pharmacovigilance, AI in Healthcare, and Drug Development.\nExperience with ML applications in clinical safety, bioinformatics, and regulatory frameworks (FDA/EMA).\nInterested in collaborating with strong deep learning / geometric modeling experts.\nHappy to contribute domain expertise, modeling strategy, and validation." } ], "comments": [ { "id": 3398843, "author_name": "PC Jimmmy", "author_url": "", "post_date": "2026-01-29T19:21:29.457000", "content": "Not sure this was a good thing. I for one stopped working this competition - fine tuning a massive model not my cup of tea when the model designed exactly for the task.
\nGuess kaggle will be able to measure if a good thing based on level of participation.
", "votes": 2, "replies": [] }, { "id": 3414446, "author_name": "Anuja Meshram", "author_url": "", "post_date": "2026-02-26T20:01:55.253000", "content": "Hi everyone,\nI’m looking to join a team.\nBackground in Pharmacovigilance, AI in Healthcare, and Drug Development.\nExperience with ML applications in clinical safety, bioinformatics, and regulatory frameworks (FDA/EMA).\nInterested in collaborating with strong deep learning / geometric modeling experts.\nHappy to contribute domain expertise, modeling strategy, and validation.
", "votes": 0, "replies": [] } ], "raw_markdown_by_id": { "3397685": "We are excited to share that the NVIDIA healthcare team have officially returned as co-organizers in Part 2 of the Stanford RNA 3D Folding Competition! \n\nAs you may have seen, research collaborators @chrismunley @youhanlee from NVIDIA BioNeMo team and KGMoN member @theoviel already posted RNAPro last week. This is the RNA structure prediction model developed by Nvidia and Kaggle hosts in collaboration with the leading Kaggler teams in the previous Part 1 of the competition, as described in our [collaborative preprint](https://www.biorxiv.org/content/10.64898/2025.12.30.696949v1). \n\nCheck out the templates and code in [Theo's post](https://www.kaggle.com/competitions/stanford-rna-3d-folding-2/discussion/668412) to get access. RNAPro, wrapped around the template-based modeling pipeline developed by @jaejohn, is also the basis of publicly shared notebooks being developed by @jaejohn @nikitakuznetsof @yanecoder and others, which are currently highly competitive on the leaderboards. Go Kaggle!\n\nWe wanted to say a huge thank you to @youhanlee, @chrismunley and @theoveil (NVIDIA) and @przemekporebski (HHMI) for all the work that was done on the Data and Validation side to get this competition launched. \n\nAs the competition contnues, you may see all of us co-organizers providing some further interesting baseline models and answering questions on forums. We are very excited to have Nvidia's continuing scientific collaboration in tackling this grand challenge in biology.", "3398843": "Not sure this was a good thing. I for one stopped working this competition - fine tuning a massive model not my cup of tea when the model designed exactly for the task.\n\nGuess kaggle will be able to measure if a good thing based on level of participation.", "3414446": "Hi everyone,\nI’m looking to join a team.\nBackground in Pharmacovigilance, AI in Healthcare, and Drug Development.\nExperience with ML applications in clinical safety, bioinformatics, and regulatory frameworks (FDA/EMA).\nInterested in collaborating with strong deep learning / geometric modeling experts.\nHappy to contribute domain expertise, modeling strategy, and validation." } }