. 2020 Dec 13;25(1):184–202. doi: 10.1111/jcmm.15903

TABLE 1.

The characteristics of included panels

Author Year	Ref^a	Participants region	Mean age (years)	HIV⁺ number	Treatment status	TB^b diagnostic method	Sample type	Original dataset	Training set	Model purpose	NcRNA^c type	NcRNA number	Modelling method	Model rebuilding
Latorre 2015	23	Southern Europe	NA^d	NA	NA	Culture	WB^e	NA	GSE29190	TB vs (HC^f and LTBI^g)	MiRNA^h	4	Linear kernel SVMⁱ	Yes
Pan 2019	15	East Asia	≥18	0	NA	Culture, smear or Xpert	PBMC^j	GSE131708	GSE131708	TB vs (HC and DC^k)	MiRNA	4	Logistic regression with forward stepwise	Yes
Wang 2011	24	East Asia	≥18	0	None	Smear or radiology	PBMC	GSE29190	GSE29190	TB vs (HC and LTBI)	MiRNA	17	SMV	Yes
Zhou 2016	25	East Asia	<18	0	NA	Comprehensive diagnosis	PBMC	NA	GSE34608 ^†	TB vs HC	MiRNA	8	Logistic regression	Yes
Barry 2018	26	East Asia	>18	0	None	Comprehensive diagnosis	Plasma	NA	GSE116542	TB vs HC	MiRNA	5	Logistic regression	Yes
Cui 2017	27	East Asia	NA	0	NA	NA	Plasma	NA	GSE116542	TB vs HC	MiRNA	3	Linear combination	No
Duffy 2018	28	South and East Africa	>18	0	Some	Culture or smear	Serum	NA	GSE116542	TB vs household contacts	MiRNA	47	Elastic‐net logistic regression	Yes
Miotto 2013‐RVM/AIC logistic regression	29	Southern Europe and east and south Africa	≥18	4	NA	Culture, smear or Xpert	Serum	NA	GSE116542	TB vs HC	MiRNA	15	RVM^l AIC^m logistic regression	Yes Yes
Qi 2012	30	East Asia	>18	0	None	Culture and smear	Serum	NA	GSE116542	TB vs HC	MiRNA	3	Logistic regression	Yes
Zhang 2013	31	East Asia	>18	0	None	Symptom, culture and radiology	Serum	SRP029907	GSE116542 ^‡	TB vs HC	MiRNA	6	Logistic regression	Yes
Alipoor 2019	32	West Asia	≥15	0	NA	Culture, smear and PCRⁿ	Exosome	NA	GSE116542	TB vs HC	MiRNA	3	Logistic regression	Yes
Hu 2019	3	East Asia	>18	0	None	Culture	Exosome	GSE116542	GSE116542	TB vs HC	MiRNA	6	Linear kernel SVM	Yes
de Araujo 2019	33	South America	>18	NA	Some	Comprehensive diagnosis	WB	GSE131174	GSE131174	TB vs (HC and LTBI)	MiRNA and snoRNA^o	4	SVM	Yes
Chen 2017	34	East Asia	≥18	0	None	Comprehensive diagnosis	Plasma	NA	GSE101805 ^§	TB vs HC	LncRNA^p	4	Logistic regression	Yes
Huang 2018	35	East Asia	>18	0	NA	Culture, smear or other aetiological evidence	PBMC	NA	GSE117563	TB vs HC	CircRNA^q	2	Logistic regression	Yes
Qian 2018	36	East Asia	>18	0	NA	Comprehensive diagnosis	PBMC	GSE103188	GSE103188	TB vs HC	CircRNA	7	Linear combination	No
Huang 2018	37	East Asia	>18	0	NA	Culture or smear	Plasma	NA	GSE106953	TB vs HC	CircRNA	2	Logistic regression	Yes
Huang 2018	38	East Asia	>18	0	NA	Culture, smear or other aetiological evidence	Plasma	NA	GSE106953	TB vs HC	CircRNA	2	Logistic regression	Yes

a: reference; b: tuberculosis; c: non‐coding RNA; d: non‐available; e: whole blood; f: healthy control; g: latent tuberculosis infection; h: micro RNA; i: support vector machine: j: peripheral blood mononuclear cell; k: disease control; l: relevance vector machine; m: Akaike information criterion; n: polymerase chain reaction; o: small nucleolar RNA; p: long non‐coding RNA; q: circular RNA.

^{^†}

The data of tuberculosis patients and healthy controls selected from GSE34608 were used as training set.

^{^‡}

Zhang et al used sequencing data of 20 samples as training set and provided the information of their training set (SRP029907); however, we only found the data of 2 samples in SRA database which was not enough to support model reconstruction. Thus, GSE116542 was used as training set.

^{^§}

The data of tuberculosis patients and healthy controls in GSE101805 were used as training set.