5 London Genomics Companies to Know About in 2026
For much of the history of genomics, the central challenge was generating sequence data quickly and affordably enough to make large-scale studies practical. As sequencing has become more accessible, the difficult questions have shifted elsewhere: which biological sequences are worth studying, how should enormous datasets be governed and analysed, what can human genetic variation tell us about disease, and how can genomic information be combined with other measurements to improve diagnosis?
That shift has broadened what a genomics company can look like, with commercial work now stretching from the generation of entirely new biological datasets to the computational and clinical problems that begin once genomic information exists.
London’s genomics sector reflects that breadth, with companies working across biodiversity, population-scale data, genetic target validation, inherited disease and epigenomics.
Here we showcase 5 companies working at different points along that range.
1. Basecamp Research: expanding the sequence space available to biological AI
Most biological AI models ultimately learn from sequence information that humans have already collected, which means their view of biology is shaped by the organisms and environments researchers have chosen to study.
London-based Basecamp Research is expanding that underlying dataset by collecting biological material from environments around the world and using metagenomic sequencing to uncover genetic diversity that is poorly represented in conventional databases.
Its BaseData resource contains more than 10 billion novel genes, with sequence data representing more than one million species not found in other genomic databases. Basecamp says its global biodiscovery network now spans more than 200 partner locations across 31 countries.
That information is used to train the company’s EDEN family of biological foundation models, which learn patterns across evolutionary sequence diversity for applications including protein design and genetic medicines.
In March 2026, Basecamp announced the Trillion Gene Atlas, an initiative developed with Anthropic, Ultima Genomics and PacBio, with NVIDIA providing AI infrastructure. The project aims to collect genomic information from more than 100 million species across thousands of sites worldwide.
Evolution has already produced an enormous library of functional proteins and genetic systems. Expanding the diversity available to biological AI widens the range of molecular solutions those models can explore.
View Basecamp Research on Benchscope →
2. Lifebit: making population-scale genomic data usable
Large genomic datasets create problems that go well beyond storage, because human genomic information is highly sensitive, often linked to clinical records and distributed across hospitals, national programmes and research institutions operating under different governance requirements.
Lifebit develops federated Trusted Research Environment infrastructure intended to make those distributed datasets accessible for approved analysis while keeping participant-level information within the environments where it is governed.
At population scale, useful analyses may require information from hundreds of thousands of participants spread across several institutions, each with its own rules governing who can access the data and what can be done with it.
Lifebit already works with organisations including Genomics England, which selected the company in July 2026 to develop the next phase of its research environment, with availability planned for 2027.
Its infrastructure gives researchers controlled access to distributed information, with the underlying records remaining under the control of the organisations that hold them.
3. Sequoia Genetics: using human variation to test drug targets
Every large human population contains millions of naturally occurring genetic differences, and some of those variants alter the activity of genes or proteins involved in disease.
That makes human genetics useful far beyond diagnosis. If a variant that reduces the activity of a particular biological target is also associated with lower disease risk, it can provide evidence about what might happen if a drug altered the same pathway. The same datasets can highlight potential safety effects or weaken the case for targets that appeared convincing in laboratory experiments.
Founded in London in 2024, Sequoia Genetics applies human genetic evidence to drug discovery and development, working with pharmaceutical and biotechnology companies, investors and academic groups.
Its work includes approaches such as genome-wide association studies, Mendelian randomisation, pQTL analysis, colocalisation and phenome-wide association studies, which can be combined to investigate the relationship between biological targets and disease.
Human populations amount to a natural experiment that has been running throughout people’s lives, long before any drug programme begins. The strength of conclusions drawn from statistical genetics depends heavily on study design and the quality of the underlying datasets, but that evidence can still influence which targets are taken forward into expensive experimental and clinical programmes.
View Sequoia Genetics on Benchscope →
4. Eye2Gene: connecting what clinicians see with the gene underneath
Inherited retinal diseases can be caused by variants in more than 270 genes, while different genetic conditions can produce overlapping appearances in the retina. Establishing the molecular cause can therefore require specialist interpretation alongside genetic testing.
London-based Eye2Gene is developing an AI platform that uses routine retinal imaging to help prioritise which genes are most likely to explain a patient’s disease, connecting visible phenotype with genomic diagnosis.
The underlying research, developed through UCL and Moorfields Eye Hospital, was published in Nature Machine Intelligence in 2025. Eye2Gene was trained on 58,030 retinal scans from 2,451 patients and externally validated using data from five additional clinical centres.
Across all test datasets, the correct gene appeared among the model’s top five predictions in 83.9% of cases for the 63 genetic causes represented by the model.
The system is designed to support genetic prioritisation, generating ranked candidate genes that can help guide subsequent testing and interpretation, especially when several genes can produce similar clinical appearances and retinal imaging provides additional evidence to narrow the search.
5. Prima Mente: reading disease through the epigenome
DNA sequence is relatively stable throughout life, but the way that sequence is regulated is much more dynamic. Chemical modifications such as DNA methylation can vary between cell types and change with ageing or disease, creating another source of biological information beyond the underlying sequence.
London-based Prima Mente is applying foundation models to this epigenomic layer in neurodegenerative disease, using cell-free DNA in blood to study biological signals that would otherwise be difficult to access directly in the living brain.
Its Pleiades models analyse DNA sequence and methylation patterns in cell-free DNA, aiming to infer where circulating fragments originated and identify patterns associated with disease.
The company is also taking this work into clinical research. SANDBOX is a UK-wide study launching across 15 NHS sites and aiming to recruit more than 1,000 patients, combining blood biomarkers, APOE genotyping and remote cognitive testing to investigate earlier and more accurate detection of cognitive decline.
Pleiades is being piloted within SANDBOX alongside those clinical and biomarker measurements.
Prima Mente is also an industry partner in the Oxford-led READ-OUT dementia diagnostics study, which is investigating how blood-based biomarkers and other diagnostic approaches could improve the way dementia is identified and monitored.
View Prima Mente on Benchscope →
What comes after sequencing?
As genomic datasets grow, interpretation matters more: connecting sequence with phenotype, establishing which associations reflect meaningful biology and deciding when the evidence is strong enough to influence research or clinical decisions.
Browse other genomics and multi-omics organisations listed on Benchscope, or explore more life science companies by location, category and focus area. If your company is listed and would like to update your profile, or if you think we have missed a relevant organisation, submit an update here.