Monocyte-related gene biomarkers for latent and active tuberculosis

Figures & data

Table 1. Clinical and demographic characteristic of subjects

Characteristics	HC	LTBI	ATB
Total number	25	10	15
Sex, male/female (% of male)	13/12(52.0)	5/5(50.0)	8/7(53.3)
Age in years
Mean	40.5	39.2	46.8
Range	21–55	25–52	23–61
Presence of cavity	0(0)	0(0)	7/8(46.7)
Bilateral disease	0(0)	0(0)	5/10(30.0)
No. (%) with positive sputum acid-fast bacilli smear	0(0)	0(0)	4/11(26.7)

HC: healthy control

LTBI: latent tuberculosis infection

ATB: active tuberculosis

Figure 1. Identification of the differentially expressed monocyte-related genes between latent and active tuberculosis. (a) The CIBERSORT algorithm was used to estimate the abundance of 22 immune cell types in the 15 samples. (b) Fraction of monocytes in PBMCs of 6 LTBI and 9 ATB. (c) The volcano plot of the DEGs in PBMCs of 6 latent- and 9 active tuberculosis PBMCs samples

Blue represents the down-regulated genes, red represents the up-regulated genes, and the green represents the differentially expressed genes with a correlation coefficient> 0.7, in monocytes.

Figure 2. GO and KEGG enrichment analyses for DEMonRGs; BP: Biological Process, CC: Cellular Component, MF: Molecular Function, KEGG: Kyoto Encyclopedia of Genes and Genomes

Figure3 The protein-protein interaction network of DEMonRGs. Green represents the down-regulated genes, and red represents the up-regulated genes. Genes in the yellow box represent hub genes

Figure 4. Expression patterns of hub DEMonRGs between latent and active tuberculosis peripheral blood samples. (a) The heat map for unsupervised hierarchical clustering of the three hub DEMonRGs in GSE54992. (b) Validation of the hub DEMonRGs in other GEO datasets

Figure 5. Verification in clinical samples. (a) SERPINA1 mRNA expression in active tuberculosis, latent tuberculosis infection and healthy controls; (b) FUCA2 mRNA expression in active tuberculosis, latent tuberculosis infection and healthy controls; (c) HP mRNA expression in active tuberculosis, latent tuberculosis infection and healthy controls; *represents P < 0.05; ** represents P < 0.01; ***represents P < 0.001; ATB: active tuberculosis, LTBI: latent tuberculosis infection, HC: healthy control. Nonparametric one-way ANOVA(Kruskal-Walls) analyses were performed on measurement of mRNA expression levels

Table 2. ROC analysis of MonRG score and random condition for discrimination between ATB and LTBI group

	Cutoff value	AUC (95% CI)	Sensitivity(%)	Specificity(%)
GSE37250
MonRG score	0.573	0.717(0.643–0.780)	56.7	79.5
Random	0.534	0.598(0.522 − 0.670)	60.8	60.2
GSE39939
MonRG score	0.463	0.769(0.648–0.863)	82.7	64.3
Random	0.343	0.545(0.417–0.668)	36.5	78.6
GSE39940
MonRG score	0.523	0.745(0.659–0.819)	70.0	72.2
Random	0.279	0.541(0.449–0.630)	74.3	38.9

AUC: Area under the curve

Figure 6. Diagnostic performance of the MonRG score for patients with active tuberculosis. (A, B, C) MonRG score between patients with active tuberculosis and latent tuberculosis in GSE37250, GSE39939, GSE39940 datasets, respectively. (D, E, F) Receiver operation characteristic curves of MonRG score for diagnosis of between active tuberculosis and latent tuberculosis infection in GSE37250, GSE39939, GSE39940 datasets, respectively. AUC: area under the curve; LTBI: latent tuberculosis infection; ATB: active tuberculosis

Data availability statement

The datasets analyzed during the present study are available from GEO database (http://www.ncbi.nlm.nih.gov/geo/). For further inquiries, please contact the corresponding author.