A hallmark of chronic inflammatory diseases is a high degree of complexity reflected by variable clinical course and disease heterogeneity. This complexity is the major obstacle standing against development of precision medicine in the field of inflammation. In a common effort, our groups achieved a major scientific breakthrough by combining deep clinical phenotyping of chronic inflam-matory skin diseases (CISD) with transcriptomics of lesional skin. This approach led to the first pre-cise disease classifier to distinguish psoriasis and eczema. With recent advances in the emerging and highly topical field of single-cell genomics, the community is starting to realize that subject heterogeneity often can be derived from heterogeneity on cell type proportions.
In this project, we aim at including this to achieve the next level, namely the development of an easy-to-use multi-disease classifier for CISD. Such a classifier is urgently needed, as diagnostics in dermatology is fallen out of time – still limited to clinical picture and descriptive histology rather than objective and precise molecular markers. To achieve our aim, we will follow two major routes that require joint and interdisciplinary efforts: 1) use state-of-the-art machine learning algorithms for feature selection and pre-diction to achieve a clinically viable classifier that is robust, highly specific and sensitive and works in routine clinical diagnostics; 2) implement single-cell sequencing to define T cell signatures in lesional skin to calibrate the system across patients. With our recently developed methodology, we will dissect T cell heterogeneity mediated by tissue pathology and use this knowledge to resolve confounding factors which then feed into the multi-disease classifier.
With our joint project, we are about to begin the era of personalized diagnostics in inflammatory skin diseases bringing forward the field in multiple ways - establishment of the first multi-classifier for CISD on a chip, new insights into pathology using single-cell transcriptomics, and the evaluation of novel biomarkers and drug targets.