Top five scientific trends shaping precision medicine in 2025
The promise of precision medicine has been intensifying over the past decade. In 2025, the field appears to be at an inflection point due to a confluence of various factors. Sequencing costs have plummeted with consistent technological development, biobanks have amassed data from millions of individuals, artificial intelligence (AI) is being embedded into discovery pipelines, gene therapy is accelerating, and regulators are evolving to meet the complexity of modern therapeutics. Below are five defining trends that are shaping the scientific landscape of precision medicine this year.
1. Next-generation genome sequencing is reshaping diagnostics and newborn screening
Clinical genomics is the starting point for personalized medicine in routine care. Diagnostic sequencing is the critical first step for most patients entering the genomics pathway.
In recent years, ultra-rapid whole-genome sequencing (WGS) has begun to transform acute and pediatric care. A landmark study published in The New England Journal of Medicine demonstrated that a cloud-distributed nanopore sequencing workflow could deliver a genetic diagnosis in just 7 hours and 18 minutes. This enabled timely and actionable diagnoses in critically ill infants and adults, informing treatments such as epilepsy management, medication selection, and even heart transplantation. In one of the reported cases of a 3-month-old infant, results from sequencing led to a diagnosis that directly impacted clinical decisions and prevented further unnecessary testing.
Building on these successes in acute care, attention is now turning to how WGS can be applied even earlier in life, with broader use of WGS for newborn screening gaining momentum. The GUARDIAN study is a population-scale initiative in New York City with a planned enrollment of 100,000 newborns. Data from the first 4,000 newborns was released and 3.7% screened positive for early-onset, actionable conditions. Notably, 110 were treatable disorders absent from standard newborn screening (such as long QT syndrome, Wilson disease, and severe immunodeficiencies). Similarly, in 2025, the UK announced plans to roll out whole-genome sequencing for all newborns within 10 years, backed by £650 million. This would dramatically expand the screening of treatable rare conditions at birth and generate a lifelong genomic baseline for individuals.
These services are rapidly being commercialized, with GeneDx releasing ultraRapid Genome in March 2025. On a recent episode of The Genetics Podcast, we discussed the importance of comprehensive newborn screening with Madhuri Hegde, CSO at Revvity, a global leader in newborn screening. She emphasized the opportunity to scale genomic testing by building on an existing process: “We’re sequencing whole genomes from dried blood spots and turning results around in under 55 hours.” This convergence of speed, accessibility, and diagnostic power can help push genomics into the front line of healthcare.
2. Biobank data is powering a new era of predictive medicine
Population genomics initiatives and national biobanks have helped build a strong foundation for precision medicine advances. Such initiatives enable researchers to uncover genetic risk factors, track disease progression, and develop more targeted therapies. These programs collect genomic data at scale, often alongside electronic health records (EHRs) and lifestyle information, to create richly phenotyped cohorts.
With the overwhelming scale and rapid execution in recent years, population genomics initiatives are beginning to power biomedical discovery. The UK Biobank (500K participants), for example, has already supported countless studies that have characterized pathogenic variants and disease biomarkers. Many such associations may go unnoticed in small cohorts.
Looking ahead, these datasets can have the strongest impact in providing high-quality data in sufficient quantities to develop accurate AI models. In line with this, a recent study showed the superior performance of a machine learning (ML) model that was trained on multi-omic data from the UK Biobank. The model predicted diseases that were still undiagnosed when participants enrolled in the program. Furthermore, it uncovered various gene-disease relationships that were previously unidentified, adding important predictive biomarkers with clinical relevance.
In parallel, AI is also enabling the translation of these large-scale research datasets into clinical-grade diagnostics. For example, SOPHiA GENETICS announced in 2025 that their AI-driven platform had analyzed over two million patient genomes. By training on diverse, real-world genomic data, the platform is now being deployed across healthcare centeres worldwide to accelerate turnaround times and improve diagnostic accuracy – demonstrating how the same scale and diversity that powers biomedical discovery can also enhance day-to-day patient care.
3. Gene therapy is scaling
A growing number of genomics-based therapies have moved into late-stage development or crossed key regulatory milestones in 2025. Many of these target ophthalmological diseases, neuromuscular diseases, metabolic diseases, and rare diseases. According to the American Society of Gene & Cell Therapy’s (ASGCT) quarterly data report from Q2 2025, there are currently 4,469 therapies in development, of which 49% are gene therapies, 29% are RNA therapies, and 22% are cell therapies. In gene therapy, 80 trials were initiated and 64% of these were in oncology.
One of the clearest examples of how far the field has come is the successful treatment of a child with a rare genetic condition that causes defects in protein metabolism. The child received bespoke CRISPR treatment that was developed in under six months, as a result of rapid collaboration across academia, industry, and regulatory agencies in the US. The infant is now growing and developing well, although it is too early to say whether the treatment is curative. This development highlights the potential and applicability of patient-specific gene editing in clinical care, in addition to the importance of collaboration and synergistic partnerships.
The FDA also granted approval, breakthrough designation, and fast track designation to various gene therapies in 2025. These ranged from treatments for familial hypercholesterolemia to geographic atrophy and Huntington’s disease. The programs span a range of delivery modalities, from lipid nanoparticles to adenoassociated virus (AAV) vectors. As these therapies progress through development, the remainder of 2025 and the start of 2026 are expected to bring a new wave of trial results and more transformative approvals, accelerating their translation to the bedside.
But 2025 has also brought a reality check. Several major players have exited or scaled back their AAV-based gene therapy programs due to safety concerns and weak commercial traction. Whether these setbacks mark a temporary reset or a long-term pivot away from AAV remains to be seen.
For a deeper dive into 2025’s gene therapy milestones, including regulatory trends and AAV market shifts, stay tuned for our upcoming State of the Genomics Industry report.
4. AI is driving gains in trial efficiency and participant matching
Patient identification and matching for precision trials can be a major bottleneck. One of the advantages of biobanks is the large-scale datasets that AI models can be trained on. Advanced algorithms are now being used to sift through patient data to identify those who meet complex inclusion criteria in a much faster and more effective manner. For instance, ConcertAI reported that its AI-powered Digital Trial Solutions platform screened oncology patients for trial eligibility more than three times faster than manual review without any loss of accuracy or added bias. Similarly, NIH’s TrialGPT tool retrieved about 90% of relevant trials for test patients while cutting clinician screening time by roughly 40%, demonstrating how large language models can streamline matching.
Beyond recruitment, these AI models, often powered by real-world data, are reshaping trial design. Researchers can simulate eligibility criteria, refine trial protocols, and identify patient subgroups at higher risk of adverse events. AI models can also generate synthetic control arms, which eliminates the need for placebo and enables more ethical trials. For example, Tempus is working with biotechs to use AI and RWD to refine trial inclusion criteria, identify high-value patient subgroups, and create propensity-matched external control arms. This can help reduce Phase II trial sizes and site requirements while strengthening the evidence base for regulatory submissions. Likewise, Unlearn.AI’s TwinRCT technology uses generative AI “digital twins” to replace part of the control arm, cutting enrollment needs by up to 50% and potentially shortening trial timelines by nearly a year.
5. A globally evolving regulatory landscape
At the FDA’s Cell and Gene Therapy Roundtable in June 2025, the agency demonstrated a willingness to adapt to meet the demands of emerging therapeutics. One of the key messages was a growing willingness to accept RWD as part of the regulatory evidence base, especially for rare diseases, bespoke gene therapies, and n-of-1 trials where traditional randomized controlled trials (RCTs) may not be feasible or ethical.
Panelists emphasized that innovative trial designs, longitudinal follow-up, and data sharing infrastructures are now essential for advancing precision therapies. Natural history studies, synthetic control arms, and platform-based approaches are increasingly being recognized as valid forms of evidence. Moreover, in response to requests for transparency around rejections, the agency later announced it would begin publicly releasing Complete Response Letters (CRLs) containing detailed explanations of why a therapy was not approved.
Still, despite these encouraging signals, many in the biotech sector remain uncertain about how consistently this flexibility will be applied. Recent leadership changes have added to the ambiguity. While public messaging emphasizes support for rare disease innovation, the current environment does not yet reflect a more certain or streamlined path forward.
In parallel, momentum is quietly shifting beyond U.S. borders. Biotech and pharma companies are increasingly looking to the UK, EU, and parts of APAC, where regulatory agencies are seen as more consistent in both communication and execution. For instance, in the UK, the government is seeking to cut trial set up times from 9 months to around 70 days. These regions are also showing strong receptivity to precision medicine approaches, making them attractive partners in development and commercialization.
The next several months are likely to demonstrate where the new administration's FDA falls on the spectrum from restrictive to permissive. The FDA has an opportunity to align its vision with day-to-day regulatory realities. Whether it does may shape where and how the next generation of precision therapies are brought to market in the US.
As a whole, the scientific foundations for precision medicine are coming into place. As diagnostics, data infrastructure, and therapies mature, 2025 is shaping up to be a year of real-world progress. What was once experimental is now becoming operational and setting the stage for broader clinical impact in the years ahead.
To learn more about the trends shaping genomics in 2025, read our full State of the Genomics Industry 2025 report.