ScaiVision is a machine learning algorithm developed by Scailyte to extract clinically relevant information from single-cell data, focusing on predictive biomarkers for therapy response and prognosis. Diana Stoycheva, Principal Scientist at Scailyte, explained that since 2017, the company has been experimenting with different single-cell data types and projects spanning various applications.
It is well known that most drugs fail in clinical trials due to a lack of efficacy. To combat this issue Scailyte are applying single cell analysis for biomarker discovery to find the right patients for the right treatment. Single cell data and omics data is usually very rich, so it provides useful information to extract these predictive biomarkers for discovery.
ScaiVision’s machine learning algorithm uses supervised representation learning to address a specific question, diagnosis or prediction of therapy to make clinical impact. To achieve this, patient data is combined with high quality clinical grade single cell data. Then the explainable machine learning model extracts clinically relevant biomarkers.
ScaiVision does not rely on clustering or previous knowledge of cell types. This makes it highly sensitive to rare cell populations and complex signatures, which is crucial for understanding complex biology and predicting therapy responses.
To demonstrate ScaiVision’s capabilities, Stoycheva presented a case study comparing standard methods with ScaiVision's approach using CAR T infusion product data. The platform identified predictive markers for neurotoxicity and long-term survival in paediatric leukaemia patients. Furthermore, they developed a diagnostic biomarker for CTCL and endometriosis using single-cell data.
Scailyte recently received a grant to generate omics data from IBD patients to find predictive biomarkers for standard treatments. In summary, ScaiVision is indication-agnostic and effective even with small patient cohorts, the company displays remarkable commitment to advancing precision medicine and collaborating on new initiatives.
