Unsupervised Learning Approach to Drive Therapeutic Precision in a Complex Disease

Unsupervised and semi-supervised Machine learning (ML) methods have a high potential for driving precision therapeutic in case of complex diseases. The diversity of its clinical phenotypes renders developing effective therapies for complex diseases a particularly challenging task. In this paper, we demonstrate how machine learning unsupervised methods can help researchers in analyzing health data in the case of Systemic Lupus Erythematous (SLE). We identified subgroups of SLE patients related to the disease severity. We analyzed the similarity between samples within these clusters and discovered distinct patterns. The clustering analysis results showed two separate patients clusters that correspond to mild and severe subgroups. The identification of well-defined subgroups of patients can facilitate the development of targeted therapeutics.