ReviveMed, Inc.

#AI-driven #Metabolomics for precision medicine l @MIT Spinout

[ReviveMed, Inc. @Revive_Med]

Generative AI has revolutionized other areas of biomedicine, and now, ReviveMed is unlocking these transformative potentials for metabolomics.
https://t.co/ytGgfJectD

[Leila Pirhaji @Leila_Pirhaji]

Starting 2026 with exciting news! Our #mzLearn work is published in Nature Communications Chemistry 🎉
LC/MS #metabolomics has huge potential for disease research, but existing methods miss real signals and detect noise. mzLearn solves this with: 🧵

📄: https://t.co/gJ1n8QKZvV

[Leila Pirhaji @Leila_Pirhaji]

mzLearn: Turning Metabolomics Data Swamps into Foundation-Model Fuel https://t.co/2JOHbXbV79

[MIT Sandbox @MITSandbox]

MIT spinout maps metabolites to uncover hidden disease drivers—reshaping diagnostics and treatment! 🧬 Read more: https://t.co/zZgm3vfHiv
#Biotech #MITStartups

[Massachusetts Institute of Technology (MIT) @MIT]

MIT spinout ReviveMed uses AI to map the body’s metabolites — such as lipids, cholesterol, and sugars — and uncover hidden drivers of disease. These insights could help researchers determine how certain treatments work and who may benefit the most. https://t.co/uyQMs9Ttfo

[Leila Pirhaji @Leila_Pirhaji]

We’re pioneering #AI-driven #metabolomics to map the body’s metabolites and uncover hidden disease drivers, opening new frontiers in precision medicine.

By leveraging #generative #AI to predict and interpret metabolic patterns, we’re unlocking deeper insights into complex diseases like cancer and cardiovascular disease, ultimately guiding more personalized and effective treatments.

From our MIT roots to building a spinout that’s making a real impact on patient care, this journey has been incredible. A huge thank you to our dedicated team, collaborators, and the MIT community for your unwavering support!

Read the full article to learn more about how we’re shaping the future of AI-powered metabolomics.

[Leila Pirhaji @Leila_Pirhaji]

Excited to share that @MIT News has featured our work at @Revive_Med!
https://t.co/YeicKRopoj

[Leila Pirhaji @Leila_Pirhaji]

I’m thrilled to announce that we’ve developed the first-in-class generative models for metabolomics powered by our innovative platform, mzLearn! 🚀

This is truly one of the most exciting times for AI research. As #generative models like DeepSeek and ChatGPT revolutionize various fields, we are now unlocking their transformative potential in #metabolomics.

With #mzLearn, a novel data-driven algorithm, we’ve automated metabolite signal detection and instrument drift correction without the need for prior knowledge or input parameters. This breakthrough has provided a robust data foundation for developing pre-trained generative AI models in metabolomics.

Using over 20,000 blood-based metabolomics profiles from diverse cohorts, we achieved:
✅ High-fidelity metabolite signal detection at an unprecedented scale
✅ Capturing metabolite representations linked to demographic and clinical variables
✅ Improved clinical predictions, outperforming clinical-grade risk scores in renal cell carcinoma (RCC) outcomes

This milestone is especially timely, as GLP-1 drugs highlight the potential of metabolic interventions in reducing risks for cancer, diabetes, and cardiovascular diseases.
📢Read our press release for more details: https://t.co/KK5vLRpcO0
📄Explore the science behind mzLearn in our preprint on bioRxiv:
https://t.co/EiXmVpCQLv

We’re thrilled to share this with the scientific community and invite collaborations to advance #precision #medicine.

[Leila Pirhaji @Leila_Pirhaji]

I am excited to share that our abstract is now live on JITC! 🚀

We've demonstrated how #AI-based metabolomic analysis complements genetic stratification, shedding light on the hashtag#interplay between #genomic alterations and #metabolism in cancer immunotherapy response. While the BMS CheckMate 025 trial linked PBRM1 mutations to immunotherapy response in renal cell carcinoma, this wasn’t reproducible in other trials. Our analysis shows that:

⭐ PBRM1 loss-of-function predicts immunotherapy response based on metabolic subtypes, independent of clinical factors like risk groups, age, and gender.
⭐ Importantly, PBRM1-mutated patients who don’t benefit from immunotherapy show significant PI3K pathway up-regulation, offering insights for future therapies.

📄 Full abstract: https://t.co/kzfzfvLLUM

Join us to learn more at #SITC!

[Leila Pirhaji @Leila_Pirhaji]

These findings help explain the inconsistent reports about PBRM1 mutations and ICI responses. They also suggest that metabolomics provides complementary information to genetics in predicting the outcomes of cancer immunotherapy.

[Leila Pirhaji @Leila_Pirhaji]

PBRM1 is the second most mutated gene in clear-cell renal cell carcinoma; however, previous clinical trials have shown conflicting results regarding its association with responses to cancer immunotherapy. Our research has discovered that patients with PBRM1 mutations respond to immune checkpoint inhibitor (ICI) therapy contingent on their metabolic subtypes.

[Leila Pirhaji @Leila_Pirhaji]

I'm excited to share that we will present at the upcoming Society for #Immunotherapy of #Cancer (SITC) conference @sitcancer, showcasing our #AI-based #metabolomics platform and findings! 👇🧵

https://t.co/lFbKwaxJDu

[Leila Pirhaji @Leila_Pirhaji]

Despite the transformative impact of cancer #immunotherapy, a significant number of patients do not respond to treatment. Our platform leverages #AI to analyze blood #metabolites to predict which patients will benefit from immunotherapy before treatment begins.

[Leila Pirhaji @Leila_Pirhaji]

I'm thrilled to share that our abstract, showcasing our AI-driven metabolomics platform, has been selected for presentation at the upcoming #ASCO conference in Chicago, May 31 - June 4.
Join us at ASCO to learn how our platform is paving the way for advancing #precision medicine

[Leila Pirhaji @Leila_Pirhaji]

We have also expanded our work in #oncology, developing predictive tests to ensure that the right patients receive the appropriate treatment. Visit our updated website for more details, and stay tuned for exciting announcements! https://t.co/WYNVInyXsv

[Leila Pirhaji @Leila_Pirhaji]

We at @Revive_Med have tackled major data challenges in detecting #metabolite signals from mass spectrometry instruments. Our #AI now identifies high-quality signals at unprecedented levels, enabling the creation of large-scale #generative AI models previously impossible.

[Leila Pirhaji @Leila_Pirhaji]

On my way to #neurips2022, excited to see talks and posters on AI in biology. DM me if you are working on GNN, deep unsurprised learning, or any AI/bio research.

[Leila Pirhaji @Leila_Pirhaji]

Such a great work, summarizing the parameters needed for LC-MS signal detection! #datadriven #metabolomics

[ReviveMed, Inc. @Revive_Med]

Congrats to our CEO, @Leila_Pirhaji, for being selected as 20 Under 40: Young Shapers of the Future (Health and Medicine)!

[MIT Sandbox @MITSandbox]

Leila Pirhaji, founder of Sandbox alumni team @Revive_Med, made Britannica's 20 Under 40 list! Take a look at the list to see her mentioned alongside some other incredible innovators here: https://t.co/ViHwiXWNro

[Leila Pirhaji @Leila_Pirhaji]

Congratulations to Prof. Vander Heiden, one of our amazing scientific advisors! @Revive_Med