2025 Life Science Trends: Insights and ModernVivo’s Vision for 2026

‍

Dear readers,

The passage of time and, more importantly, the efforts of scientists around the world have brought about considerable change in the field of drug discovery and in the policy that governs our work. It can be a lot to keep up with, so we thought it would be helpful to share our year in review: a ModernVivo take on what’s happened in 2025 and where we’re headed in 2026. Read on to learn more about:

In this newsletter, we’ll discuss important shifts in:

• Science & Research — new drug targets, modalities, and delivery pathways.
• Market & Policy — business and regulatory changes that influence development timelines, cost structure, CRO demand, and capital flow.
• Artificial Intelligence — new tools and expanded AI capabilities for drug discovery

If you have ideas or thoughts on what we should include in future blogs, please reach out! The team at ModernVivo loves hearing about the topics that spark your interest and curiosity. Onwards!

Scientific Discovery

Across biology, chemistry, and translational pharmacology, in 2025 we’ve seen breakthroughs in how drugs function and which diseases can be targeted. We’re entering an era where previously “undruggable” targets are now being reached by new drug modalities. Here are a few stories that stood out to us.

Protein-targeting small molecules are being used as molecular gates and protein degraders

Gate Bioscience, a well-funded startup based in Brisbane, California, announced a collaboration and license agreement with Eli Lilly & Company to discover, develop, and commercialize so-called “gate therapeutics” — a new class of small molecule drugs that eliminate disease-causing proteins by blocking their transit out of the cell in which they were made.

The total deal value could reach $856M and may capitalize on the advantages of small molecule drugs, which tend to have better oral bioavailability and manufacturability, while enabling functionality closer to that of biologics.

According to a September report by the American Chemical Society on recent trends in drug discovery, Proteolysis Targeting Chimeras (PROTACs) are also continuing to make waves, with over 80 PROTAC drugs in the development pipeline and over 100 commercial organizations involved with ongoing PROTAC development. Cancer remains the leading disease in PROTAC-related literature.

These small molecules operate via an induced-proximity modality that drives protein degradation by bringing together the target protein with an E3 ligase, such as cereblon, VHL, MDM2, or IAP. Scientists are working to identify new ligases and unlock insights into their structure and function in order to target various proteins previously thought of as inaccessible. Advancements in PROTACs may lead to fewer off-target drug effects.

Click here to watch an ACS webinar on drug discovery trends and breakthroughs of 2025.

Gene writing and editing, and CAR-T therapies are moving in vivo

This past year has also brought very interesting advancements in in vivo CAR-T therapy and in vivo gene editing. This means the preclinical landscape in 2026 and beyond may favor more transformative biologic, cell, and gene-based drug modalities, reshaping therapeutic and investment categories in the coming years.

And the money is already moving. In August, Gilead entered into a definitive agreement to acquire Interius BioTherapeutics for $350M. The acquisition was intended to strengthen Gilead’s position in the cell therapy market by integrating Interius’s in vivo platform for the generation of CAR-T cells directly within a patient’s body. This may yield therapies with more lasting effects by inserting DNA into patients’ genomes directly and eliminate traditional steps in CAR-T therapies like cell harvesting, engineering, and reinfusion.

Just a few months later, Regeneron Pharmaceuticals announced a $275M partnership deal with Tessera Therapeutics to develop and commercialize a rare disease in vivo gene writing program, known as TSRA-196. This one-time treatment for a rare genetic disorder, alpha-1 antitrypsin deficiency (AATD), has shown preclinical potential in editing the gene which encodes the alpha-1 antitrypsin protein, whose inactivity is directly responsible for AATD’s respiratory and hepatic disease pathology.

With each of these new advancements comes the need for regulatory frameworks to ensure safety and efficacy, especially if we move to manufacture these therapeutics at scale. The translational implications of all of this new science are massive as we head into 2026 and beyond.

Market & Policy

Beyond scientific advancements, 2025 was defined by some major structural shifts in the business and regulatory environment.

Contract Research Organizations are in demand but also facing stiff competition

ModernVivo has been keeping a close eye on the preclinical contract research organization (CRO) market. As of 2025, the CRO market continues to expand, buoyed by a growing number of emerging biotech firms and a rising pipeline of cell and gene therapies. Emerging biotechs, often resource-constrained and managing multiple preclinical assets, are increasingly outsourcing to CROs that offer flexibility, niche expertise, and scalable operations. Meanwhile, complexity of modern therapeutics and regulatory demands are putting pressure on traditional full-service CROs, making room for specialized CROs and hybrid-service models. While publicly available data does not yet show a collapse of legacy CROs, the structural signals suggest a potential re-segmentation of the CRO market in favor of specialized players.

The FDA signals a move away from mandatory animal testing

In April, the Food and Drug Administration (FDA) published their Roadmap to Reducing Animal Testing in Preclinical Safety Studies. The document outlines a “strategic, stepwise approach for FDA to reduce animal testing in preclinical safety studies with scientifically validated new approach methodologies (NAMs), such as organ-on-chip systems, computational modelling, and advanced in vitro assays.” We believe this will have profound implications for cost, timelines, predictive accuracy, and R&D strategy.

The document emphasizes that traditional animal testing is scientifically weak with >90% of drugs that appear safe and effective in animals failing in humans due to unexpected issues. The roadmap frames NAMs (such as in vitro, in silico, and microphysiological systems) as more predictive of human biology, faster, and cheaper than animal testing.

These tools already outperform many animal models (e.g., Liver-Chip predicting 87% of hepatotoxic drugs) and reduce reliance on costly non-human primates, speeding up development and improving early decision-making. With FDA policy now enabling animal-study waivers and encouraging NAM-based submissions, companies will need to redesign preclinical strategies around human-first data, computational modeling, and regulatory readiness for NAM integration.

It’s important to note that these changes in the FDA are echoing earlier guidance put forth in the European Union, such as the EU roadmap for phasing out animal testing for chemical safety assessments, and the European Medicines Agency’s non-clinical NAMs expert community platform for information sharing on the reduction of animal testing.

The US Senate and House Negotiate Details of the BIOSECURE Act

A 2025 revision of the BIOSECURE Act is being reintroduced as an amendment to the National Defense Authoritization Act (NDAA), which may restrict the use of “biotechnology equipment or services” from certain “biotechnology companies of concern (BCCs),” by US federal agencies, or by entities receiving US federal grants or contracts. Unlike earlier versions, the 2025 draft does not explicitly name entities as BCCs (e.g. BGI, MGI, Complete Genomics, WuXi AppTec, and WuXi Biologics). Instead, it establishes that any entity included in the annual list published by the US Department of Defense as a Chinese military company operating in the United States (known as the “1260H List”) qualifies as a BCC.

In October, the Senate approved the revised version of BIOSECURE as part of the “must-pass” NDAA legislation. However, the version of the NDAA passed by the House did not include BIOSECURE, meaning the bill must still go through a reconciliation process between the House and Senate before it becomes law. 

Nevertheless, industry stakeholders can expect some version of BIOSECURE to be retained and that some form of the legislation will become law by early 2026. For biopharma companies, the BIOSECURE Act signals an imminent shift toward trusted, transparent, non-adversarial supply chains, meaning firms must audit their CDMO/vendors now, reduce reliance on at-risk foreign biotech providers, and proactively build compliant, manufacturing and data-handling systems to avoid future federal-funding or contracting disruptions.

Artificial Intelligence

A ModernVivo year in review would not be complete without a look into how AI is changing the game of drug discovery. Namely, 2025 demonstrated just how fast AI is transforming from simple tools that accelerate parts of the drug development pipeline to full co-developers that can coordinate entire drug development workflows. Here’s what we noticed.

LLM-based multi-agent system simulates the full drug discovery workflow

Researchers at Tsinghua University’s Institute for AI Industry Research published a paper which demonstrates how LLM-driven agents equipped with specialized machine learning models and computational tools can identify potential therapeutic targets, discover promising lead compounds, enhance binding affinity & key molecular properties, and perform in silico analyses of toxicity and synthetic feasibility.

They named the system PharmAgents and the results are compelling. For instance, PharmAgents identifies biologically relevant targets with minimal error accumulation. In a sample task, 16 of 18 predicted targets were confirmed appropriate by human experts. The LLMs also correctly mapped diseases to validated therapeutic targets, including JAK1/JAK2/JAK3 and IL-4Ra for atopic dermatitis. When it came to lead identification & optimization, PharmAgents significantly improved binding affinity, drug-likeness, and synthesizability compared to other models. They cite a success rate improvement from 15.7% to 37.9%.

While the findings are nascent, the implications for preclinical drug discovery and R&D are major. PharmAgents is the first system to demonstrate end-to-end capability across tasks traditionally requiring interdisciplinary human teams. Further, accurate metabolism and toxicity predictions and feasible synthetic pathways mean LLMs can eliminate dead-end molecules earlier, lowering cost and time of preclinical work. 

And we’re already seeing this idea enter the market with companies like Lila Sciences which is building “the world’s first scientific superintelligence platform and autonomous labs for life, chemical, and materials sciences,” with the aim of letting AI design experiments, run them, observe results, and iterate. Since emerging from stealth, Lila has attracted major investor interest to the tune of about $550M. This level of funding and institutional support suggests that the vision of fully automated, AI-driven preclinical discovery may soon become reality.

Diffusion models can generate 3D molecular graphs

Chemistry, Machine Learning, and Biomedical Engineering researchers at Carnegie Mellon University developed a pharmacophore-conditioned diffusion model for 3D molecular generation. Their model, named PharmaDiff, shows that diffusion models conditioned solely on 3D pharmacophores can generate diverse, novel, drug-like molecules that match pharmacophore constraints with high fidelity and outperform some structure-based models in docking. This research leads the way for rational drug design for targets lacking structural data, one of the biggest bottlenecks in hit discovery.

Anthropic and Google are following in ModernVivo’s footsteps

In October, Anthropic announced Claude for Life Sciences, a specialized version of their Claude AI model, optimized specifically for life sciences research and drug discovery. While their aim is less specialized than ModernVivo’s, they are taking a step in the right direction by supporting researchers with many key steps in preclinical R&D workflows, such as hypothesis generation, drafting protocols & documentation, data analysis & bioinformatics workflows, and regulatory & compliance support. Their new tool also features various integrations that link Claude directly to other widely used life science tools.

In November, Google also released Google Scholar Labs, an experimental AI extension of Google Scholar meant to help users answer more complex research questions. While the tool is still in limited-access beta, it has the potential to speed up biomedical and preclinical literature review. Sound familiar?

See you in 2026!

ModernVivo has been spending a lot of time thinking about how we can build tools to give preclinical R&D teams confidence in study reproducibility. We’ve been diving deep into the wonderful world of bioinformatics in order to assess how we might leverage large data sets to create reproducibility scores associated with the peer-reviewed studies ModernVivo delivers. Stay tuned for updates about awesome new ModernVivo product features in 2026! As always, if you have ideas or feedback on how we may continue to improve, drop us a line. Your insights are a gift.

From the ModernVivo team to yours, we wish you a very Happy Holidays! May you find rest and inspiration in time off with your loved ones. We’ll see you on the other side of 2025. Until then, keep doing great work.

Warm regards,
ModernVivo
Seattle, WA, USA