2025's Top 5 Drug Discovery Highlights And How To Stay Ahead In 2026
By Ray Dogum, Chief Editor, Drug Discovery Online
Editor’s note: Thank you to all our loyal readers who enjoy our original content and gain valuable insights from our experienced drug discovery expert community. Your support keeps this show going.
When I stepped into the Chief Editor role halfway through 2025, I began a deep dive across the drug discovery landscape. This included an assessment of the regulatory environment, funding gaps in basic science, global R&D competition, and AI breakthroughs empowering researchers.
Below is a concise summary of the year’s most consequential shifts affecting this rapidly evolving field. I couldn’t cover all bases as it was such a standout year for biotech. I’m always eager to hear perspectives from our readers so let me know what I missed.
1. NAMs Move Center Stage to Bolster Preclinical Toolkits
The FDA’s mAbs-first roadmap (published in April) to reduce animal testing sparked an industry-wide rush to understand, test, and develop New Approach Methodologies (NAMs). Subsequent announcements from both industry and regulatory leaders have reshaped early discovery workflows as sponsors leaned into organoids, tissue chips, and in silico models to flag potential drug risks (and opportunities) earlier, with the aim of compressing development timelines.
The NIH announced it will no longer award funding to new grant proposals that rely exclusively on animal testing, effective July 2025. All new funding opportunities must include NAMs or other human-based research approaches.
The NIH’s new Office of Research Innovation, Validation and Application (ORIVA) office amplified funding for human-relevant assays, while the EMA’s horizon scanning report clarified context-of-use expectations for NAMs in preclinical packages. The U.K. published its robust policy paper on replacing animals in science, further validating the importance of NAMs.
From an industry perspective, Charles River Labs, a global leader in preclinical research, has built its reputation on animal-based testing for decades. Their recent push into NAMs (including in vitro, in silico, and organ-on-chip technologies), marks one of 2025’s significant inflection points toward mass NAMs adoption. When a CRO of this scale shifts from traditional animal models toward human-relevant alternatives, it sends a powerful signal to the entire drug development ecosystem: NAMs are no longer niche, they’re becoming mainstream.
Drug Discovery Online’s (DDO’s) take: Collaborate with NAMs developers and (when it makes sense) use these tools for early tox and DMPK. Engage precompetitive pilots (e.g., FNIH VQN) and publish benchmarking datasets to build community trust in NAMs.
Replacing the long-standing reliance on ‘gold standard’ mouse studies with NAMs will demand not only technical innovation but a fundamental cultural shift across research and regulatory ecosystems. Let’s not forget: over 90% of drugs which pass all animal models are not approved by the FDA. Check out a recent DDO panel discussion to learn more about the factors impacting NAMs adoption.
2. Industrial-Scale AI Redefine The Drug Discovery Process
2025 marked AI’s graduation from point solutions to platform-scale engines for discovery. For target identification workflows, omics-driven insights were paired with AlphaFold3’s expanded structural predictions, which now cover DNA/RNA, ligands, and antibody-antigen complexes. This leap gave researchers unprecedented clarity on binding modes and selectivity hypotheses.
We saw OpenFold3 release its open-source alternative to AlphaFold3, offering commercially usable co-folding capabilities for protein-ligand and antibody-antigen complexes. Through the AI Structural Biology (AISB) Network, pharma companies began fine-tuning AI models with proprietary structural data via federated learning, creating a path to pharma-grade models without sharing raw IP.
Hit ID and lead optimization workflows saw equally dramatic progress. Boltz-2, an open-source biomolecular foundation model from MIT and Recursion, delivered near physics-level accuracy for binding affinity predictions at speeds up to 1,000× faster than traditional free-energy perturbation simulations, making large-scale virtual screening practical for early-stage teams. Meanwhile, Chai Discovery’s Chai-2, backed by OpenAI and Anthropic, redefined biologics discovery with zero-shot antibody design, achieving 16–20% hit rates (a 100x improvement over previous computational benchmarks).
Lilly partnered with NVIDIA to launch TuneLab, an AI-driven platform designed to train and deploy models end-to-end across the drug discovery and development continuum. The initiative also allows selective access for biotech companies, enabling them to leverage certain models without building full-scale infrastructure.
This year saw the highest single-year jump in IND filings for AI-originated molecules, driven by companies like Insilico Medicine, Recursion, BenevolentAI, Absci, and Generate Biomedicines. Most filings were small molecules concentrated in oncology, fibrosis, autoimmune disorders, and rare diseases, reflecting AI’s strength in target discovery and generative chemistry. Published in Nature Communications, Insilico Medicine used its Chemistry42 platform to design a first-in-class PROTAC with a dual-action mechanism targeting PKMYT1, a notoriously challenging cancer target. This demonstrated AI’s ability to tackle complex modalities, not just small molecules.
DDO’s take: AI is no longer optional for competitive discovery. Invest in data provenance, model governance, and wet-lab feedback loops (the trifecta for credible AI-driven drug design). And watch for disruptors like OpenFold3, Boltz-2, and Chai-2, which democratize advanced modeling and enable bespoke optimization strategies at scale.
3. Clinical Success Refocus Drug Discovery On Personalized Treatments
In May, Children’s Hospital of Philadelphia and Penn Medicine delivered the world’s first personalized in vivo therapy to an infant with CPS1 deficiency, a urea cycle disorder driven by a single pathogenic variant. Baby KJ was treated with a personalized CRISPR base-editing therapy delivered via lipid nanoparticles to correct his CPS1 gene mutation in liver cells. The case encourages policy on “n=1” pathways, CMC agility, and ethics. KJ Muldoon was featured on Nature’s “10 People Who Helped Shape Science” as the Trailblazing baby and he continues to hit developmental milestones.
The first personalized CRISPR therapy surely made headlines, but its real impact for discovery is upstream: it proves variant-to-therapy pipelines can be compressed when bioinformatics, modular editing platforms, and rapid CMC converge.
Additionally, personalized mRNA cancer vaccines, like mRNA-4157 (V940) in combination with pembrolizumab, showed a 44% reduction in recurrence risk for advanced melanoma (Phase 2b). Another big win for the RNA industry and the “n=1” research pioneers who are leading those developments.
DDO’s Take: Expect rising demand for target validation frameworks that accommodate rare disease variants, and preclinical assays that confirm edit specificity and durability before clinical translation.
4. China Is A Formidable Early-Stage Innovator
In 2025, China cemented its position as a global powerhouse in early-stage drug innovation. Five of the year’s ten largest R&D licensing deals originated from Chinese companies, accounting for 38% of all big-pharma in-licenses with $50M+ upfront payments. These deals weren’t limited to traditional assets. Two focused on AI-driven discovery platforms (AstraZeneca–CSPC Pharma and the $6B Dovetree–XtalPi deal), while others centered on bispecific and multispecific antibodies, underscoring China’s strength in next-generation biologics. GSK’s $12B agreement with Hengrui Pharma, granted global rights to a PDE3/4 inhibitor for COPD and 11 additional candidates spanning respiratory, immunology, and oncology.
This surge reflects a decade of regulatory reform and capital inflow that transformed China from a generics hub into an innovation engine. Domestic pipelines now feature first-in-class small molecules, ADCs, PROTACs, and gene therapies, with multinationals establishing strategic research centers in Beijing to tap this momentum.
DDO’s Take: Expect China to double down on AI-integrated biologics, expand global co-development partnerships, and push for regulatory harmonization with the FDA and EMA to accelerate cross-border research efforts.
5. Early-stage Biotechs Bear the Brunt of 2025’s Funding Volatility
The funding environment in 2025 proved especially unforgiving for early-stage biotechs. While overall biopharma venture investment reached $5.8 billion across 86 rounds in Q3, this represented a decline from $6.6 billion in the same quarter of 2024, and the year is on track to post the lowest IPO activity since 2010.
Capital deployment increasingly favored later-stage programs. Median round sizes for Phase II companies climbed to $63 million, up from $50 million in 2024, while Series B and later rounds captured $11.1 billion across 158 deals year-to-date, widening the gap with Seed and Series A activity. This selective approach reflects heightened risk aversion amid persistent macroeconomic headwinds and elevated interest rates, which pushed investors toward assets with clearer clinical and commercial pathways.
For early-stage ventures, the squeeze was twofold: fewer transactions and smaller checks.
For founders, 2025 underscored the importance of robust data packages and strategic partnerships to weather a market that rewarded maturity over promise.
The historic 43-day U.S. government shutdown compounded the pain, freezing NIH grant reviews, delaying BARDA and ARPA-H proposals, and stalling SBIR/STTR funding. NIH budget cuts deepened uncertainty, forcing startups to slow hiring and potentially shelve early programs.
The result: early-stage biotechs faced longer fundraising cycles, tougher valuation negotiations, and an increasingly competitive race to demonstrate differentiated science.
DDO’s Take: Persistence drives success. Securing funding takes more than compelling data and a good narrative. Investors demand clarity and credibility. They look for detailed execution plans, verified CRO quotes, and a team of proven experts ready to deliver measurable results. Consider alternative deals with non-profits and, if you’re brave, decentralized science organizations to help expand your pool of potential investors.