AI-driven drug development has taken a major step forward in the decades-long fight against HIV, as a newly unveiled biotech company makes its public debut with a significant scientific breakthrough.
Viva-Thera, incubated by XTALPI (02228), announced that it has achieved a structural breakthrough in the MPER region of HIV, leveraging XtalPi's AI-powered drug discovery capabilities to accelerate the nomination of a preclinical candidate compound (PCC) in just seven months.
Using AI-driven virtual screening and molecular design, the project has not only reached PCC nomination but is also advancing four distinct HIV drug development pipelines simultaneously, all powered by XtalPi's AI platform. This achievement targets a market where a single drug can generate over ten billion US dollars in annual sales, while approximately 41 million people worldwide living with HIV continue to require ongoing treatment.
Where the breakthrough lies
Viva-Thera, founded by structural biology expert Dr. Fu Qingshan, is making its first public appearance with XtalPi serving as both incubator, strategic shareholder, and core technology partner. The newly disclosed progress demonstrates the combined ability of both entities to transform scientific discoveries into viable drug candidates.
The seven-month timeline to PCC nomination reflects a dual breakthrough combining biological understanding with AI-driven molecular design and validation. This speed is built on complementary strengths: Viva-Thera contributes rare structural and target insights, while XtalPi integrates these into its AI screening, molecular generation, and experimental iteration systems. The PCC nomination proves that this collaboration can now produce concrete drug development candidates.
Why MPER matters
The MPER region and its adjacent transmembrane domain play a critical role in HIV's entry into human cells, with approximately 90% sequence conservation across different viral strains. For the highly variable HIV virus, MPER represents an ideal targeting direction, but the region's proximity to the viral membrane and its complex structure have historically limited precision design efforts.
Viva-Thera has now resolved the atomic-level structure of the MPER-TMD-CT domain for the first time, providing XtalPi's AI with a more precise design foundation. The combination of structural information for this challenging target, experimentally confirmed antiviral activity, and PCC nomination together justify continued development of this candidate.
From virtual screening to PCC
XtalPi's AI first screened a virtual library of tens of millions of compounds, identifying over ten specific binding candidates. Subsequent experiments confirmed that five of these exhibited strong viral inhibitory activity. Building on these results, both teams employed AI-driven molecular generation and structural optimization to finalize the PCC drug molecule.
The entire process from initial screening to PCC nomination was completed in just seven months. As an industry benchmark, the traditional timeline from drug discovery to the preclinical stage averages approximately 4.5 years. Furthermore, XtalPi's algorithms reduced computational consumption during virtual screening by 95% compared to conventional methods. For drug development that requires continuous screening, validation, and optimization, lower computational investment and shorter candidate discovery cycles mean teams can obtain critical results earlier and make quicker decisions on resource allocation.
Expanding into vaccine development
The same structural foundation is also supporting vaccine research. A novel MPER immunogen developed by Viva-Thera induced antibody levels approximately five times higher than traditional HIV immunogens in rhesus macaques. Even when antiserum was diluted 1,000-fold, it maintained strong inhibitory activity against HIV pseudovirus infection. The first metric demonstrates immune response intensity, while the second reflects the actual antiviral function of antibodies. Together, these results indicate that structural design has translated into observable biological effects.
Based on the conserved nature of the target, the team proposes that this candidate vaccine could potentially cover over 90% of circulating viral strains, further expanding the scope of broad-spectrum prevention research. These findings respectively support the development potential of both small-molecule drugs and vaccines.
Market potential and unmet needs
The market has already provided clear commercial benchmarks for effective HIV innovation. In 2025, Gilead's HIV treatment drug Biktarvy generated $14.3 billion in annual sales, a 7% year-over-year increase. A single product has achieved a ten-billion-dollar annual revenue stream. Long-acting innovative products also demonstrate significant commercial value. The HIV prevention drug Yeztugo, administered once every six months, carried a U.S. annual list price of $28,218 at launch. This price does not account for insurance adjustments, discounts, or assistance programs, nor does it represent global pricing, but it clearly illustrates the pricing scale for long-acting HIV innovations.
Meanwhile, unmet needs remain substantial: approximately 41 million people worldwide were living with HIV in 2025, with about 1.2 million new infections that year. The treatment side requires more convenient, durable, and resistance-resistant options, while the prevention side needs products that are more accessible and capable of sustained protection. Existing product sales and pricing demonstrate the considerable commercial capacity of HIV innovation, while long-term medication burdens and accessibility issues leave room for new products to improve upon.
Four pipelines targeting different needs
Viva-Thera's long-acting small molecules and broad-spectrum vaccines correspond directly to these identified needs. Should the company establish advantages in efficacy, dosing frequency, or cost, it stands to enter this global market. The company is also advancing toward deeper curative goals.
Beyond small molecules and vaccines, Viva-Thera is developing a gene delivery system targeting CD4+ T cells, utilizing gene editing tools to address HIV proviral DNA integrated into the host genome. The team has completed vector construction, cell infection experiments, and relevant humanized mouse implantation work, while also advancing in-situ CAR-T and universal CD4+ T cell programs in parallel.
Additionally, the team is collaborating deeply with Jietai Technology (07666.HK) and its AI-driven nano-delivery platform NanoForge to jointly advance HIV mRNA vaccine development. From XtalPi's AI molecular design to Jietai's AI nano-delivery, the technical synergy within XtalPi's incubation ecosystem is forming a more complete R&D chain, accelerating industrial translation and value creation.
Each of the four pipelines carries a distinct mission: vaccines prevent infection, small molecules block viral entry, and gene and cell therapies target the root cause of persistent infection. The viral reservoir, in particular, is the core reason existing treatments cannot achieve a cure and why viral rebound occurs after treatment cessation—making it a critical focus of curative research.
A shared AI-driven foundation
All four pipelines share the same underlying AI-powered R&D system from XtalPi: computation guides design, and experimental feedback drives optimization. The seven-month PCC nomination for this small-molecule program provides concrete evidence of the system's ability to accelerate candidate drug discovery.
As a strategic shareholder, XtalPi holds the rights to participate in Viva-Thera's future value growth, with its multimodal AI drug discovery platform potentially helping Viva-Thera unlock a multi-tens-of-billion-dollar HIV treatment market. For XtalPi, this achievement also serves as a compelling validation of its AI4S platform: when AI combines with deep scientific discovery, the process of advancing challenging targets into concrete drug candidates is accelerating, and the platform's commercial potential is rapidly converting into verifiable pipeline assets.
Comments