Analysis of disease-state specific variance in the peripheral blood of rheumatoid arthritis patients

Abstract

Rheumatoid Arthritis (RA) is an autoimmune disease afflicting people of all ages and sexes and has been studied for decades. In the past, bulk RNA-sequencing has been used to study the disease, which does not account for heterogeneity in cell populations. In this study, we present early findings from the first single-cell RNA sequencing study of the peripheral blood of RA patients. Through this study, we attempt to understand disease-specific variation in cell-type proportions and transcriptomic signatures. We also use this study as a means to compare the performance of two scRNA-seq feature selection algorithms, DUBStepR and HVG, a critical part of the single-cell analysis pipeline. Due to the lack of controls, an external control dataset was added to the in-house-generated patient data, with batch effect correction. Our analysis revealed that both DUBStepR and HVG had comparable performance, allowed good separation of cell types present and had similar patient proportions in each cluster. Our analysis determined that two of the patients had a more severe disease state compared to the rest of the cohort, which was not reflected in the clinical scores. Thus, this study also provides preliminary evidence that transcriptomic profiles may contain relevant information to aid in patient stratification.