Abstract
BACKGROUND: To determine single-cell-type gene expression in peripheral blood (PB) requires either prior cell sorting or single-cell RNA sequencing. We developed a novel ratio-based biomarker (RBB) called Direct Leukocyte Subpopulation-Transcript Abundance (DIRECT LS-TA) that allows quantification of monocyte-specific gene expression directly from PB without cell sorting. METHODS: DIRECT LS-TA leverages proportional cell counts and differential gene expression profiles among leukocyte subpopulations to identify monocyte-informative genes. Using a new ICEBERG plot (Figure 1) based on a mathematical model of cell-mixture gene expression, we shortlisted genes with 2.5-fold higher expression in isolated monocytes compared to PB, indicating > 50% of transcript contribution by monocytes alone. PSAP and CTSS were identified as monocyte informative reference genes with low biological variation. Using one of them as the denominator, another monocyte informative target gene is used as the numerator to derive the RBB. The method was validated for detection of host response towards bacterial infection across multiple datasets. FINDINGS: Over 50 monocyte-informative genes were identified, including immune response genes such as VNN1, IL1B, NLRC4 and IFI44L. DIRECT LS-TA results showed excellent correlation with gold standard isolated monocyte expression (R2 = 0.55-0.97). VNN1 RBB showed consistent upregulation across five datasets (median 2.7-fold, P < 10-8) with good diagnostic performance (AUC = 0.84-0.99). Other genes including NLRC4, CYP1B1 and NFKBIZ were also useful biomarkers. CONCLUSION: DIRECT LS-TA provides a reliable way of quantification of monocyte-specific gene expression from PB without the need of cell sorting and demonstrated potential use for rapid infection detection and antibiotic stewardship.