Biostate AI and Weill Cornell Medicine Collaborate to Develop AI Models for Personalized Leukemia Care
HOUSTON & NEW YORK--(BUSINESS WIRE)-- Biostate AI, a leading innovator in artificial intelligence for RNA sequencing, and Weill Cornell Medicine have entered into a strategic collaboration to develop AI for personalized assessments of patient disease prognosis and evolution. Initially, the collaboration will focus on leukemia, leveraging the Weill Cornell Leukemia Program's vast biorepository of bone marrow and blood samples.
Under the agreement, Biostate AI will leverage its proprietary barcode-integrated reverse transcription (BIRT) technology for RNA sequencing (RNAseq) to analyze RNA expression from patient samples at scale. These results will be used to pre-train and fine-tune a transformer-based AI model, using RNA expression as its "language." Together, Biostate AI and Weill Cornell Medicine will look to prototype and then validate an AI model that will be able to perform acute myeloid leukemia (AML) subtype stratification, disease prognosis, and therapy selection.
'Current personalized medicine approaches look only at one or a few key biomarkers for each disease,' said David Zhang, co-founder and CEO of Biostate AI, and former Associate Professor of Bioengineering at Rice University. 'This results in poor predictive accuracy because every patient is different, and we are not looking at the whole picture of a patient's health or disease. We aim to revolutionize patient treatment via precise AI-driven decisions on whether to prescribe treatments such as bone marrow transplants and BCL-2 inhibitors. This may ultimately lead to patients living longer and healthier lives."
Under the partnership, the Biostate AI team will initially work with Dr. Gail J. Roboz, Professor of Medicine at Weill Cornell Medicine. In the pilot phase of the project, the teams will develop a prototype AML-specific AI model using 1000 retrospective bone marrow and blood samples. Upon reaching pre-determined technical performance milestones, the teams will expand the study to include up to 50,000 retrospective samples, including multiple longitudinal samples from the same patient, to develop definitive models.
"AML is an aggressive, biologically heterogeneous malignancy that poses both diagnostic and therapeutic challenges. My team and I look forward to investigating the potential for Biostate AI's approach to help personalize treatment selection for AML patients,' said Dr. Roboz.
Cornell University's Center for Technology Licensing (CTL) , facilitated the agreement, in line with the Weill Cornell Medicine's commitment to translational research and innovation in precision medicine.
'This agreement reflects Cornell's commitment to fostering industry partnerships that translate academic research into real-world impact,' said Dr. Lisa Placanica, Senior Managing Director of CTL at Weill Cornell, a part of Weill Cornell Medicine Enterprise Innovation, which is dedicated to nurturing innovations and industry collaborations for the academic medical institution.
Since its founding in 2023, Biostate AI has invented and patented over 12 technologies in RNAseq and AI, and has partnered with more than 100 different research groups across the United States using its BIRT technology.
'Disease prognosis represents an important new direction for Biostate AI, and unlocks applications never before possible with AI as a field,' said Ashwin Gopinath, co-founder and CTO of Biostate AI, and former Assistant Professor of Mechanical Engineering at MIT. 'Traditional RNA biomarker discovery has led to the development of standard-of-care tests like OncotypeDx for breast cancer, but this approach has failed to generalize to other diseases or even to other cancers. By taking on a modern AI plus massive data approach, we believe that Biostate AI is developing a general-purpose technology possible for improving outcomes for patients suffering from all diseases.'
About Biostate AI. Biostate AI is a startup building generative AI that predicts the evolution of human disease and drug response based on RNA sequencing data. Its patented wetlab technologies, including BIRT, allow affordable and scalable collection of massive amounts of transcriptomic and genomic data. With sites in Houston, TX, Palo Alto, CA, Bangalore, India, and Shanghai, China, Biostate AI is an international company with collaborations with top hospitals, academic researchers, and biotech/biopharma companies. Biostate AI was founded by serial entrepreneurs and former professors David Zhang (Rice U.) and Ashwin Gopinath (MIT) and is backed by top investors and experts including Accel, Dario Amodei (Anthropic), and Mike Schnall-Levin (10X Genomics).
Hashtags
Try Our AI Features
Explore what Daily8 AI can do for you:
Comments
No comments yet...
Related Articles
Yahoo
16 hours ago
- Yahoo
New coronavirus discovered in China ‘only small step' from infecting humans
A new coronavirus discovered in China is only a small step from mutating and causing another global pandemic, experts have warned. Scientists believe the variant, called HKU5-CoV-2, may infect a broader range of animals than Covid-19 – which caused millions of deaths – and may have more potential for jumping between species. US researchers fear that HKU5-CoV-2, found in China, in February, could also infect humans, leading to a widespread outbreak. The new study, published in Nature Communications, looked at a lesser-known group of coronaviruses called merbecoviruses, which includes HKU5 and MERS-CoV, which is responsible for the deadly Middle East Respiratory Syndrome. The team from Washington State University looked at how the new pathogen interacts with human cells. They found that a small change in the virus's spike protein could allow it to attach to human ACE2 cells in people's throats, mouths and noses. HKU5-CoV-2 can infect and replicate inside human cells in both the airways and gut. According to the World Health Organisation, about 35 per cent of people infected with Middle East Respiratory Syndrome die. Since 2012, some 27 countries have reported cases, leading to 858 known deaths due to the infection, which spread from camels. But when HKU5 was discovered in February, scientists warned against exaggerating the risks because it does not enter human cells as readily as Sars-CoV-2, which caused Covid-19. HKU5 was first detected in bats by scientists from the Chinese laboratory where some say Covid originated in 2019. Prof Michael Letko, a virologist who co-led the study, said: 'HKU5 viruses in particular really hadn't been looked at much, but our study shows how these viruses infect cells. 'What we also found is HKU5 viruses may be only a small step away from being able to spill over into humans.' When Covid-19 emerged it was widely blamed on markets in China where different breeds of wild animal are kept caged and often slaughtered close to other animals. Meat is sold at the open-air stalls. Critics said the markets were the perfect breeding ground for new zoonotic diseases – those that spread to humans – to emerge. The scientists, whose experiments studied how the new pathogen interacts with human cells, believe the virus would have to carry certain mutations if it were to infect humans. 'These viruses are closely related to MERS, so we have to be concerned if they ever infect humans,' Prof Letko said. 'While there's no evidence they've crossed into people yet, the potential is there and that makes them worth watching.'
Yahoo
20 hours ago
- Yahoo
TSA Announces Big Change at Seattle-Tacoma International Airport
Over the past several months the Transportation Security Administration has been rolling out new technology to screen passengers before their travels. One of the busiest airports in the United States will be seeing that upgrade this month. On June 5 the TSA announced a new security checkpoint located at the south end of the terminal on the arrivals level at Seattle-Tacoma International Airport. The new checkpoint is designed to screen up to 750 passengers per hour. "It will be open daily 4 a.m. to 8 p.m. This is the first checkpoint on the baggage claim level as part of innovative designs in a space constrained terminal. This will provide a new option for travelers for convenience as well as additional queuing and re-composure space to improve the customer experience," the TSA said in a statement. The security administration revealed what passengers can expect when encountering the latest version of Credential Authentication Technology (CAT-2). "A TSA officer will scan the passenger's photo identification and a camera will capture a real-time photo of the passenger. CAT-2 uses facial matching technology to compare the features on the photo ID against the in-person, real-time photo," the TSA said. "Once the unit confirms a match, a TSA officer verifies it and the traveler can proceed to security screening. TSA officers can perform additional passenger verification if needed. Through a secure Internet connection, the units also verify that an individual is ticketed for air travel, negating the need to show a boarding pass." The TSA also revealed photos captured by the new equipment are "never stored or used for any other purpose than immediate identity verification." Travelers who don't want to participate in the facial matching process have the option to opt Announces Big Change at Seattle-Tacoma International Airport first appeared on Men's Journal on Jun 6, 2025
Yahoo
a day ago
- Yahoo
Nvidia Stock (NVDA) Preserves Pack Leader Status Following Q1
Nvidia (NVDA) once again proved to the markets, after reporting its Q1 earnings, that it remains the undisputed leader powering the global AI revolution, driven by relentless demand for its chips, even amid ongoing geopolitical and trade headwinds. Easily unpack a company's performance with TipRanks' new KPI Data for smart investment decisions Receive undervalued, market resilient stocks right to your inbox with TipRanks' Smart Value Newsletter Beyond beating all key metrics (excluding one-off events) and offering guidance that resonated well with investors, Nvidia stock experienced a strong post-earnings surge, even if the momentum cooled slightly in the days that followed. That said, there's arguably still a missing spark needed to fully reignite the stock's momentum heading into 2025. However, considering the broader growth story, Nvidia continues to trade at a very attractive valuation—one that could deliver meaningful alpha over the long term. Short- to mid-term bumps, especially tied to macro risks in China, are worth monitoring but don't alter the core thesis. Given the company's strong execution and still-intact fundamentals, I continue to rate NVDA as a Buy. As I pointed out in a previous article, for Nvidia stock to perform well after its Q1 Fiscal 2026 results, it wouldn't be enough to simply beat estimates—it needed to crush them and deliver guidance that topped market expectations. And that's precisely what Nvidia did. The company reported revenue of $44 billion (as shown in the chart below), beating its own guidance of $43 billion—a massive 69% increase year-over-year. Gross margins, which had been a point of concern during the early stages of Blackwell's rollout, came in at 71.3% (excluding the H20 charge, a financial write-off tied to its China-specific H20 GPUs). That's also above the guided range of 70.6% to 71%. Even more impressive was the Q2 guidance. Nvidia is projecting $45 billion in revenue, with a margin of plus or minus 2%. At the high end, that's $45.9 billion—above the ~$45.5 billion consensus estimate leading into earnings day. Margins are expected to rise again, landing between 71.8% and 72.0%, with a margin of error of approximately 50 basis points. At the top of that range, it suggests another round of margin expansion, which is exactly what investors wanted to hear. The company continues to state that once Blackwell production is fully ramped, margins could move into the 70–80% range over the next few quarters. This stronger-than-expected performance, along with the recovery in margins, led to long-term EPS estimates getting bumped up by around 9% starting in FY2029. Revenue projections for those years were also revised higher by roughly 8%. Not surprisingly, Nvidia shares jumped more than 5% in after-hours trading following the earnings release, although the stock mostly leveled off in the sessions since. In my view, the muted post-earnings reaction is actually a positive sign—it shows the market didn't see enough red flags to justify a selloff in Nvidia stock. However, bears argue that even with revenue jumping nearly 70% year-over-year, signs of fatigue in the growth story are evident, with sequential growth of just 12% potentially indicating that Nvidia's explosive momentum is entering a more seasonal or plateauing phase. For a company priced for hypergrowth, this kind of quarter-over-quarter slowdown can be an early warning signal. And it's not just about the numbers—geopolitical and regulatory pressures are starting to have real consequences. The U.S. export restrictions on AI chips like the H20 led to a $4.5 billion write-down and forced Nvidia to walk away from an estimated $15 billion in potential sales to China. That's not just a short-term financial hit—it also opens the door for competitors like Huawei to gain ground, especially as they accelerate domestic chip development. Although analysts have raised long-term estimates following Q1, there remains a genuine possibility that Nvidia's global dominance could face challenges over time, particularly if policy pressures persist. When factoring in the impact of the H20 write-off, Q1 would have been the first quarter since the AI boom began in which Nvidia did not achieve sequential growth. And for a stock valued for perfection, even a modest slowdown can pose a valuation risk. To me, the bigger issue here is that this goes beyond Nvidia—it's about the strategic direction of U.S. tech leadership. As CEO, Jensen Huang has warned that if the U.S. continues down this restrictive path without a more balanced strategy, it may ultimately strengthen Huawei and erode America's edge in AI. That's the kind of long-term headwind that bears are likely to latch onto as the growth narrative gets more complicated. But setting the macro risks aside, Nvidia's bull case remains intact after Q1, especially when considering its growth. The company continues to stand out as a GARP (growth at a reasonable price) opportunity. Currently, Nvidia trades at 31.6x forward earnings, with a consensus growth rate of 29% CAGR over the next three to five years. This results in a PEG ratio of just 1x—virtually identical to Advanced Micro Devices (AMD), despite AMD's significantly slower growth outlook, and significantly lower than most of the other Magnificent 7 names. Of course, for that valuation to hold up, Nvidia's growth trajectory needs to remain clean and strong. But honestly, it's hard to think of another company with Nvidia's size and scale, this level of fundamental quality, and such massive exposure to secular tailwinds like AI, trading at such an attractive growth-adjusted valuation. Among the 40 analysts who've covered NVDA in the past three months, there's hardly any room for doubt: 35 rate the stock as a Buy, while only four suggest Hold. Not a single analyst rates NVDA stock as a Sell. Currently, NVDA's average stock price target is $173.57, implying a potential upside of 22% from the current share price. Setting aside one-off events, Nvidia left the bears with almost nothing to complain about—crushing its own guidance and delivering a Q2 forecast that dispels doubts about how quickly gross margins are recovering. While risks like evolving U.S. sanctions on China, Nvidia's ability to maintain its presence in that market, and rising local competition are worth keeping an eye on, they don't outweigh the current risk-reward of going long, especially given the company's strong growth trajectory. Disclaimer & DisclosureReport an Issue Error in retrieving data Sign in to access your portfolio Error in retrieving data Error in retrieving data Error in retrieving data Error in retrieving data
