Our vision is to provide patients and clinicians with the benefits of early diagnosis of a broad spectrum of diseases by combining the power of protein design and machine learning.
Rosa was spun out from the University of Bristol in 2019 based on >10 years of research in the lab of world-leading protein designer Prof. Dek Woolfson. We are backed by experienced business angels including the founders of Ziylo, the biosensing company recently acquired by Novo Nordisk for up to $800 million, and Cramer, the systems technology company acquired by Amdocs for $425M.
Since incorporation, Rosa's multidisciplinary team has demonstrated that its innovative sensing platform, Pandra, can detect life-threatening diseases with high accuracy in patient samples. We are currently collaborating with scientists, clinicians and industry to further develop and validate Pandra, and bring our first tests into the clinic.
We are a proud part of the Science Creates deep tech ecosystem in the heart of Bristol. We operate a dedicated laboratory at Science Creates St Philips Incubator, which is home to over 20 deep tech companies and provides a melting pot of academic researchers, startup and spinout companies and investors.
Pandra is Rosa's biosensing platform. It combines arrays of designed barrel-forming peptides with a back-end cloud database and machine learning algorithms to create a powerful biosensor for detecting disease using patient samples.
Arrays of designed peptide barrels
Rosa’s arrays of peptides simultaneously interact with many small molecule biomarkers of disease in patient samples.
These interactions cause differential displacement of bound fluorescent dyes, creating a fluorescence fingerprint for a sample.
Patient diagnosis by machine learning
Groups of fingerprints are used to train machine learning algorithms, which classify unknown samples to provide a diagnosis.
Pandra has advantages over competing single-biomarker sensors:
Pandra is not reliant on pre-identified biomarkers of disease, making it ideally suited when traditional biomarkers are unavailable, non-specific or difficult to detect.
Disease-specific optimisation by peptide design, screening and testing.
Low-cost & easily deployable plate reader format for deployment early in the care pathway.
Andrew R Thomson, Christopher W Wood, Antony J Burton, Gail J Bartlett, Richard B Sessions, R Leo Brady, Dek N Woolfson
Franziska Thomas, William M Dawson, Eric J M Lang, Antony J Burton, Gail J Bartlett, Guto G Rhys, Adrian J Mulholland, Dek N Woolfson
ACS Synthetic Biology, 2018
William M Dawson, Freddie J O Martin, Guto G Rhys, Kathryn L Shelley, R Leo Brady, Dek N Woolfson
Chemical Science, 2021
We are currently using Pandra to detect several chronic diseases that present significant healthcare challenges and clinical needs.
We leverage our peptide design expertise to optimise arrays for specific diagnostic challenges. To develop a new sensor, we require only annotated samples to screen our designed peptide library and train our models. This means Pandra is widely applicable across a range of applications.
To find out more about our platform and how it can aid clinical decision making in your area of interest, get in touch below.