Fig. 8: Gene signature-based classification of macaque brainstem RNA-seq samples and public malaria samples.

The MetaIntegrator R package⁹⁹ was used to compute the meta-scores of the macaque brainstem RNA-seq samples (representing biological replicates n = 4 Pcoat⁻ MB⁻, n = 5 Pcoat⁺ MB⁻ and n = 9 Pcoat⁺ MB⁺) based on a public CM gene signature: BCL2L13, NFIX, S100P, S100A8, RETN, PGLYRP1, BIRC5, IL1R2, MCOLN1, MED25, FUCA1, LZTFL1, SLC25A38, RPA1, RPIA, ACVR1, ATP6V0E2, SPSB3, PIGQ⁵⁶, and b the nine identified CM biomarkers. The meta-scores were displayed as violin plots grouping samples by infection/treatment (right panels). P-values comparing group means on the violin plots were computed using Wilcoxon rank sum test (two-sided). The meta-scores were used for classifying samples as associated with cases or controls and the detection threshold was varied to generate a receiver-operating characteristics (ROC) curve (left panels). c The nine identified biomarkers were used with MetaIntegrator R package to analyze the public gene expression datasets where human subjects were infected with P. falciparum. Violin plot of the meta-scores of n = 16 UM, n = 26 SMA, n = 131 CM and n = 17 healthy controls derived from public microarray datasets GSE1124, GSE33811, GSE72058, GSE116306 and GSE117613, computed using the set of nine candidate biomarker genes. P-values comparing group means on the violin plots were computed using Wilcoxon rank sum test (two-sided test, unadjusted p-values comparing each condition against CM).