Blood protein analysis could help diagnose rare diseases missed by genome sequencing

Researchers from Queen Mary University of London, the Berlin Institute of Health at Charité (BIH) and Genomics England have shown that measuring proteins in the blood can provide important additional clues about the effects of genetic variants, helping to identify diagnoses and potential new disease-causing genes that genome sequencing alone has been unable to resolve.

The study analysed blood samples from people with rare diseases who remained without a genetic diagnosis following analysis through Genomics England’s 100,000 Genomes Project.

By measuring nearly 1,500 proteins in blood and combining this information with genomic data, the researchers were able to resolve previously uncertain genetic findings and provide evidence supporting diagnoses for some patients.  The approach also helped identify candidate genetic variants and genes that could potentially explain the conditions of other patients and warrant further investigation.

The findings provide proof-of-principle that large-scale blood proteomics – the measurement of many proteins simultaneously – could complement genome sequencing in the diagnosis of rare disease.

Rare diseases collectively affect millions of people worldwide, but finding the genetic cause of an individual patient’s condition can be extremely challenging. Although genome and exome sequencing have transformed rare disease diagnosis, a large proportion of patients still receive no definitive answer.

One particular challenge is interpreting genetic changes known as “variants of uncertain significance” (VUS). These are variants identified through sequencing where there is not enough evidence to determine whether they are responsible for a patient’s disease.

Proteins offer an additional source of information because they can provide a functional readout of what is happening as a result of a genetic variant. An unusually high or low level of a particular protein in a patient’s blood can therefore provide evidence that a genetic change is having a biological effect.

Dr Julia Carrasco-Zanini, of Queen Mary University of London’s Precision Healthcare University Research Institute and first author of the study, said:

“Genome sequencing has transformed our ability to diagnose rare diseases, but for many patients it still doesn’t provide an answer.

“Our study shows how looking at proteins alongside the genome can give us another layer of evidence. If a genetic variant is accompanied by an unusually low level of the corresponding protein, for example, that can help us understand whether that variant is actually disrupting the way the gene functions.

“This could help us make more sense of genetic information we already have, rather than simply looking for more and more variants.”

The researchers found the approach was particularly useful in patients with hereditary haemorrhagic telangiectasia (HHT), a rare inherited disorder affecting blood vessels.  Combining protein and genomic information helped resolve cases that had remained undiagnosed following genomic analysis.

Protein measurements also helped researchers identify genetic changes that had been missed by standard approaches to analysing genome sequencing data.

The team additionally investigated whether unusual protein levels could point towards genes not previously firmly established as causing a particular rare disease.

One example involved TIE1, a protein involved in blood vessel function. Researchers identified a rare genetic variant in TIE1 in a patient with a previously unexplained inherited cardiac disorder who also had exceptionally low levels of the TIE1 protein in their blood.

The same genetic variant was present in the patient’s father, who had the same condition, and was not found in other participants in the 100,000 Genomes Project. Further laboratory experiments using cells derived from the patient supported the finding by showing markedly reduced levels and signalling of the TIE1 protein.

The researchers stress that findings such as this represent candidate gene-disease links requiring further evidence rather than confirmed new causes of disease. However, they demonstrate how combining protein and genomic information could help researchers prioritise promising leads that would otherwise be difficult to identify.

Professor Damian Smedley, of Queen Mary University of London and a corresponding author on the study, said:

“For people living with a rare disease, reaching a diagnosis can be a long and frustrating process. Genome sequencing has been enormously powerful, but inevitably there are cases where the genetic information alone isn’t enough.

“What is exciting about this approach is that protein measurements can provide another piece of the puzzle. They can help us decide which genetic changes are most likely to matter and where it may be worth going back to the genome for a much more targeted search.”

Unlike some existing approaches for studying the effects of genetic variants, which may require skin biopsies and the growth of patient cells in the laboratory, the researchers used blood samples. This could potentially offer a more scalable and less burdensome way of obtaining additional biological information from patients.

The researchers emphasise that the current technology measures only a proportion of the proteins encoded by the human genome, and not every disease-relevant protein can be reliably detected in blood.  In addition, not all disease-causing genetic variants will alter the amount of a protein circulating in the bloodstream.

This means the approach cannot yet provide answers for every undiagnosed patient.  But as proteomic technologies become more comprehensive and sensitive, the findings suggest they could become an increasingly valuable complement to genomic analysis.

Professor Claudia Langenberg, Director of Queen Mary University of London’s Precision Healthcare University Research Institute, Head of the Computational Medicine group at BIH and a corresponding author on the study, said:

“The real opportunity is in bringing different layers of biological information together. The genome gives us the blueprint, but proteins can tell us something about how that blueprint is being translated into biology in an individual patient.

“This study demonstrates the potential of combining these technologies to extract more information from genomic data and, ultimately, to improve our ability to find answers for people affected by rare diseases.”

The researchers say larger studies across a wider range of rare diseases and more diverse populations will now be needed. Improvements in the number of proteins that can be measured, assay sensitivity and the development of robust reference ranges will also be important before this type of approach can be incorporated more widely into clinical diagnosis.

The research brought together scientists and clinicians from Queen Mary University of London, the Berlin Institute of Health at Charité – Universitätsmedizin Berlin, the Max Delbrück Center for Molecular Medicine and Genomics England.

Athanasios Kousathanas, Principal Genomics Data Scientist at Genomics England and a corresponding author on the study, said:

“One of the biggest challenges in rare disease genomics is understanding which genetic variants may be contributing to a person’s condition. Protein data can provide additional evidence to help researchers investigate the biological consequences of genetic variation.

“We were pleased to support this collaborative study, which explored how proteomic and genomic data can be analysed together. The findings provide early evidence that combining different types of biological data may help researchers extract additional insights from existing genomic datasets.

“Further research will be needed to understand where this approach could be most useful, but studies such as this are an important part of exploring how genomic data can be used alongside other sources of biological information to improve our understanding of rare diseases".

Further information

Julia Carrasco-Zanini et al.: Proteomics identify disease-associated variants in patients with rare diseases undiagnosed after genome sequencing. Sci. Transl. Med. 18 (2026). DOI: 10.1126/scitranslmed.aeb1331