Rosa publishes founding research in Nature Communications
We are delighted to publish the founding research behind Rosa and our Pandra platform in Nature Communications. The paper - a collaboration between the research group of our co-founder, Professor Dek Woolfson (a global research leader in protein design) and Rosa Biotech - is a major milestone on the companies journey to bring Pandra to patients in the clinic.
The paper describes how the extraordinary diversity of protein structures designed by the Woolfson group over the last decade were used to produce biosensors applicable in medical diagnostics. It centres around the arrays of synthetic receptor proteins (called alpha-helical barrels), which are central to the Pandra technology. The design of these structures has previously been the subject of publications in Science, Nature Chemistry and Nature Communications.
The research team designed an array of 46 of these receptors, which (when bound to a fluorescent dye) could detect subtle changes in sample composition through interactions with small metabolites (or biomarkers) present in the samples. In patient blood, many of these metabolites are important indicators of disease progression, and the team predicted that their sensor could be used to detect these changes, and potentially diagnose disease.
The sensor works through a method called differential sensing. First, the sensor array is used to analyse lots of serum (blood) samples from patients with a known disease, and lots of samples from a control group (e.g., healthy individuals). Second, the collected fingerprints are used to train a machine learning model to recognise the difference between the 'patient' fingerprint and the 'healthy' fingerprint. Finally, this trained model can be used to predict whether an unknown sample comes from a patient with the disease, or a healthy individual. This forms the basis for the Pandra platform, which is depicted in the diagram below.
In comparison to traditional sensing methodologies that dominate current clinical practice, this approach is hypothesis-free, meaning no prior knowledge of the target biomarkers is required to train the sensor. This means it can detect diseases where sensitive and specific biomarkers are either unknown or not easily detectable.
The authors demonstrate that this array can be used to differentiate serum samples from healthy individuals from those with non-alcoholic steatohepatitis (NASH), a type of chronic liver disease. NASH is extremely prevalent, affecting roughly 1 in 20 adults globally, but is currently diagnosed at a late stage in most cases. Therefore, new diagnostics that can help detect NASH earlier are a critical area of unmet need and clinical interest. Rosa is currently working to validate these initial findings in a larger cohort of patients and to bring Pandra into the clinic. We hope to be able to share details of these studies in the coming months.
Congratulations and thanks to all involved, including Prof Dek Woolfson, Dr Will Dawson, Dr Kathryn Shelley, Dr Chris Wood, Dr Guto Rhys, Dr Lucia Lombardi and the rest of the team across the Bristol BioDesign Institute, Bristol Chemistry and Biochemistry Departments.
Read the full paper here.