Fig. 8

Linear regression analysis demonstrates the IFN signature is most closely related to monocyte cell surface transcripts. Linear regression analysis using SLEDAI values from the patients of 5 SLE WB and 2 SLE PBMC datasets and the patient GSVA enrichment scores for cell type specific signatures (Supplemental Table 12). a Cell types or signatures with significant non-zero slopes (p < .05) related to SLEDAI by linear regression analysis in at least half of the datasets which had determinable GSVA scores (Supplementary Data 13) were used to determine overall significance of the regression lines and the r² predictive values for all 7 SLE datasets with available SLEDAI information (Supplementary Data 14, 15). Cell type or process GSVA enrichment categories with linear regression p values < .05 are shown and r² predictive values are listed after the cell type or process. b Representative plot using the HepC-IFNA2 signature for the linear regression analysis between the IFN signature with overlapping transcripts to the cell type or process signatures removed (Supplementary Data 16) and the cell type or process GSVA enrichment score for the patients from 10 SLE WB and PBMC datasets (Supplementary Data 17, 18). Cell types or signatures significantly (p < .05) related to HepC-IFNA2 score in at least half of the datasets which had determinable GSVA scores were used to determine overall regression lines for all 10 datasets. r² predictive values are listed after the GSVA enrichment category. Supplementary Data 19–28 show relationships and linear regression analysis for the other IFN signatures. c For time-course dataset GSE72747, linear regression analysis was done for the change in the core IFN GSVA scores versus the change in monocyte cell surface scores between 0 and 12 weeks and 12 and 24 weeks