Gilead Sciences, Inc.

RIYADH, SAUDI ARABIA, October 29, 2025 / EINPresswire.com / -- On day three of the Global Health Exhibition 2025, King Faisal Specialist Hospital and Research Centre (KFSHRC) signed a series of agreements at its pavilion, underscoring its commitment to high-impact partnerships that drive innovation, medical education, and biotherapeutics, aligned with its vision to be the provider of choice in specialized care. To strengthen its role as a national reference center for biotherapies, KFSHRC signed an agreement with Roche to improve spending efficiency and stabilize supply chains for high-value biologic medicines, giving patients local access to advanced treatments that meet the highest quality standards. A framework agreement with Kite will expand the availability of advanced immunotherapies within KFSHRC facilities, reducing referrals abroad and improving cost efficiency. KFSHRC also signed an MoU with US-based CellCo to accelerate AI- and bioengineering-enabled therapies, including a clinical study involving heparin, the first biologic anticoagulant, alongside collaboration on novel treatments for cancer, immune conditions, and chronic neurological disorders. An MoU with US-based Kopra Bio aims to speed the development and localization of manufacturing for innovative therapies and highlights the intent to explore a value-based partnership model that includes in-kind contributions from KFSHRC in exchange for equity participation at Kopra Bio. In genomics, KFSHRC entered a framework agreement with Al-Jeel Medical to deploy advanced gene-sequencing in diagnostics and research. The work includes specialized panels in oncology, rare diseases, and pharmacogenomics to pinpoint disease-causing genetic factors and support personalized treatment plans that improve outcomes and quality of life. KFSHRC also signed an MoU with Novartis covering clinical studies, patient support programs, continuing medical education, and innovative therapies. The agreement includes establishing a Center of Excellence for radioligand therapy and advanced nuclear medicine to transfer know-how and localize cutting-edge technologies for difficult-to-treat cancers, expanding access to the latest global treatments within the Kingdom. To build local talent, an MoU with LEORON will deliver innovative training and professional development programs that embed global best practices and strengthen the competitiveness of Saudi healthcare professionals. KFSHRC has been ranked first in the Middle East and North Africa and fifteenth globally among the world’s top 250 academic medical centers for 2025 and recognized by Brand Finance as the region’s most valuable healthcare brand. It is also listed among Newsweek’s World’s Best Hospitals 2025, Best Smart Hospitals 2026, and Best Specialized Hospitals 2026, reaffirming its leadership in innovation-driven care. Riyadh King Faisal Specialist Hospital and Research Centre email us here Legal Disclaimer: EIN Presswire provides this news content "as is" without warranty of any kind. We do not accept any responsibility or liability for the accuracy, content, images, videos, licenses, completeness, legality, or reliability of the information contained in this article. If you have any complaints or copyright issues related to this article, kindly contact the author above.

The AI-focused startup Genesis Therapeutics has a new name — and a new AI model it believes has best-in-class performance. In an interview with Endpoints News , CEO Evan Feinberg said the new model, called Pearl (short for Placing Every Atom in the Right Location), outperformed Isomorphic Labs’ AlphaFold 3, Chai Discovery’s Chai-1, and other top AI models across several protein-ligand benchmarks. Feinberg called predicting these structures “one of the most notoriously challenging problems to solve.” AlphaFold 3 has been seen as a leader in this space since its May 2024 debut . “For the problem of modeling protein-ligand interactions, we’re confident that Pearl is the best model that exists today,” Feinberg said. (To be sure, the AlphaFold 3 release is now over a year old. Isomorphic is a “significant generation or two ahead internally in terms of what our models can do, so think like 4, 4.5, 5,” Isomorphic’s chief AI officer Max Jaderberg told Endpoints earlier this month .) The Bay Area biotech has also changed its name to Genesis Molecular AI. Founded in 2019, Genesis has raised over $300 million, including a $200 million Series B in 2023 . Feinberg was formerly a graduate student in Vijay Pande’s lab at Stanford, where he developed physics-based machine learning models that turned into Genesis. The biotech now has roughly 130 employees and is advancing its own preclinical pipeline with no disclosed timeline to entering the clinic, as well as partnerships with Gilead and Incyte. Genesis detailed the claims about its model in a preprint posted Tuesday from more than 30 authors, including some Nvidia engineers. The paper used three metrics, including “Runs N’ Poses,” an academic-created test which found in February that models like AlphaFold 3 appear to largely memorize their training data. That raised the question of whether these models have seen enough data to actually learn and generalize. Genesis’ Pearl achieved an 85% success rate on the Runs N’ Poses test, compared to 74% for AlphaFold 3, 74% for Boltz-1x, and 70% for Chai-1. The team also tested how these models fared against a higher bar of accuracy, defining success as predictions less than an angstrom away from the actual structure. (The standard test, on which Pearl scored 85%, looks for predictions within two angstroms.) Against that tougher metric, Pearl had a 70% success rate, compared to 62% with AlphaFold 3, 57% with Boltz-1x, and 56% with Chai-1. The release of the preprint and model coincides with Nvidia’s AI conference in Washington, DC. Feinberg credited Pearl’s outperformance to Genesis’ long-running focus on integrating more physics knowledge into its models. That included taking a page from how Waymo developed its autonomous cars using synthetic data. Synthetic data are derived from experiments using physics and allowed Genesis’ models to see more examples, particularly in data-scarce areas, like small molecules in the Protein Data Bank. “We are the first group to show any evidence of scaling law phenomena here, where we can generate more synthetic data with physics, pretrain the model with this data, get better performance and repeat,” Feinberg said, describing this idea as still being in the “early innings.” The MIT research team behind Boltz has already extended its modeling to other protein design tasks, like predicting binding affinity and, most recently, generating new protein binders . Genesis’ preprint does not include Pearl’s performance in predicting binding affinity or potency, and Feinberg said that was beyond the scope of the paper. Having high-accuracy structures is a crucial first step to then tackling predictions like potency, Feinberg said. “A lot of these co-folding models hallucinate, just like any other diffusion model or LLM,” Feinberg said. “They will make something that prima facie looks OK, but any expert looks at it, and is like, ‘That’s not quite right.’” Editor’s note: A previous version of this story stated Pearl was compared to Chai-2. The story has been corrected to reflect Genesis compared its model to Chai-1.

Plus, news about Cybin, Incyte, Roche, Merck, Eisai, Neuphoria Therapeutics, Aldeyra Therapeutics and X4 Pharmaceuticals: 🏭 FDA issues CRL to Regeneron: The drugmaker has officially been handed a complete response letter for its pre-filled syringe for Eylea HD. The company had warned in August that it expected regulatory delays after issues were found at a Catalent site in Bloomington, IN, that’s now owned by Novo Nordisk. The plant has since been issued an “official action indicated” classification. Another customer of the Indiana site, Scholar Rock, also received a CRL related to the plant. — Anna Brown 🧪 AbbVie scraps asset licensed from Dragonfly Therapeutics: The Chicago-area pharma giant terminated a Phase 1 test of the first asset that it licensed from Dragonfly . AbbVie started testing ABBV-303/DF4101 in 2024 in solid tumors. Dragonfly CEO Bill Haney told Endpoints News that the company’s other programs with AbbVie are ongoing. Dragonfly has multiple internal candidates in clinical testing as well as Phase 1 programs with Bristol Myers Squibb and Merck. It also has a preclinical pact with Gilead. — Kyle LaHucik 💵 Cybin’s $175M direct offering: The company is selling about 22.3 million shares at $6.51 apiece from several blue-chip investors, including Venrock Healthcare Capital Partners, OrbiMed and Point72. It’s developing a psychedelic called CYB003 as a treatment for major depressive disorder. — Jaimy Lee 🛠️ Incyte halts work on two drugs: Under new CEO Bill Meury, the drugmaker has paused development of its anti-CD122 candidate, called INCA034460, and its BET inhibitor, called INCB57643. The biotech also ended povorcitinib’s development in chronic spontaneous urticaria and hinted that more pipeline adjustments are in the works. “We are taking a deliberate approach to pipeline prioritization. We are actively reviewing our R&D efforts and focusing on high-value programs that are scientifically differentiated, address unmet medical needs, and have the potential to significantly drive Incyte’s next phase of growth,” Meury said in a release.  Incyte also raised its revenue guidance for the year. It’s now in the range of $4.23 billion to $4.32 billion. — Kyle LaHucik 🔬 Roche’s Gazyva succeeds in kidney disorder: The drug met its primary endpoint in a Phase 3 study for idiopathic nephrotic syndrome, a rare autoimmune disease that affects the kidneys. Roche will head to the FDA to discuss next steps for the drug, but the timing is not yet clear. Gazyva was approved earlier this month for the treatment of active lupus nephritis. — Max Gelman 💊 Merck, Eisai tout kidney cancer data: The combination of Merck’s Welireg and Eisai’s Lenvima achieved a statistically significant improvement in progression-free survival over Exelixis’ Cabometyx in advanced renal cell carcinoma. Overall survival did not reach statistical significance, though Merck and Eisai noted this was an interim analysis. There was also a “trend toward improvement” on OS. — Max Gelman 📈 Neuphoria Therapeutics adopts poison pill after Phase 3 fail: The biotech’s board selected the shareholder rights plan to “protect the interest of the company and its stockholders.” The move lowers the chances that a person or group will accumulate enough shares to control Neuphoria. The biotech’s social anxiety drug failed a Phase 3 trial last week, and the company decided to drop the asset, named BNC210. The biotech’s stock price $NEUP was up almost 6% in premarket trading on Tuesday. — Kyle LaHucik 💼 Aldeyra Therapeutics updates its pipeline: The biotech discontinued work on ADX-629 in patients with Sjögren-Larsson Syndrome. It also said it expects to file INDs for a pair of RASP modulators sometime next year. — Jaimy Lee 💰 X4 Pharmaceuticals ended up raising $155.3 million in its public offering, up from the $135 million it set out to raise. — Jaimy Lee