Prediction of Deep Vein Thrombosis Based on GC-MS and Machine Learning

doi:10.7503/cjcu20240199

Abstract

Abstract:

This study investigated the changes of endogenous metabolites in the serum of deep vein thrombosis（DVT） rats， screened characteristic metabolites related to DVT， and constructed the prediction models for clinical diagnosis and forensic identification of DVT. The DVT rat model was constructed using the inferior vena cava ligation method， and blood samples were collected 72 h post-surgery. Gas chromatography-mass spectrometry（GC-MS） was used to analyze the small molecule metabolism profile in the rat serum. Then， orthogonal partial least squares discriminant analysis combined with the Mann-Whitney U test initially identified 22 differential metabolites associated with DVT， involving metabolic pathways， such as glyoxylate and dicarboxylate metabolism， tricarboxylic acid cycle（TCA cycle）， and alanine， aspartate and glutamate metabolism. Subsequently， a feature selection method based on the random forest classification algorithm（Boruta） was applied to screening out 13 characteristic metabolites correlated with DVT from the differential metabolites and predictive models for DVT were constructed using 3 machine learning algorithms（logistic regression， linear discriminant analysis and Adaboost ensemble learning algorithm）. The results showed that the DVT prediction model of linear discriminant analysis has a high performance with an accuracy of 87%， as well as precision of 0.88， recall of 0.86， F1 score of 0.87， and area under the receiver operating characteristic curve（AUROC） of 0.95. The research indicates that the DVT prediction model constructed using GC-MS metabolomics combined with machine learning algorithms can provide technical support for the diagnosis， treatment， and forensic identification of DVT.

Key words: Metabolomics, Deep vein thrombosis, Machine learning, Feature selection, GC-MS

CLC Number:

O657

TrendMD:

HOU Zejin, LI Rongqi, LI Jian, FENG Yining, JIN Qianqian, SUN Junhong, CAO Jie. Prediction of Deep Vein Thrombosis Based on GC-MS and Machine Learning[J]. Chem. J. Chinese Universities, 2024, 45(9): 20240199.

Figures/Tables 8

References 23

1	Raskob G. E.， Angchaisuksiri P.， Blanco A. N.， Buller H.， Gallus A.， Hunt B. J.， Hylek E. M.， Kakkar A.， Konstantinides S. V.， McCumber M.， Ozaki Y.， Wendelboe A.， Weitz J. I.， Arterioscler. Thromb. Vasc. Biol.， 2014， 34（11）， 2363—2371
2	Liu Y. L.， Zhao Y. J.， Zhang G. H.， Chinese J. Forensic Medicine， 2016， 31（4）， 424—426
	刘玉利，赵一杰，张国徽. 中国法医学杂志， 2016， 31（4）， 424—426
3	Sung Y.， Spagou K.， Kafeza M.， Kyriakides M.， Dharmarajah B.， Shalhoub J.， Diaz J. A.， Wakefield T. W.， Holmes E.， Davies A. H.， Eur. J. Vasc. Endovasc. Surg.， 2018， 55（5）， 703—713
4	Li X. Q.， Zhang F. X.， Wang S. M.， Chinese Journal of Vascular Surgery（Electronic Version）， 2017， 9（4）， 250—257
	李晓强，张福先，王深明. 中国血管外科杂志（电子版）， 2017， 9（4）， 250—257
5	Di Nisio M.， Squizzato A.， Rutjes A. W.， Büller H. R.， Zwinderman A. H.， Bossuyt P. M.， J. Thromb. Haemost.， 2007， 5（2）， 296—304
6	Stroup B. M.， Nair N.， Murali S. G.， Broniowska K.， Rohr F.， Levy H. L.， Ney D. M.， J. Nutr.， 2018， 148（2）， 194—201
7	Taegtmeyer H.， Circulation， 2016， 134（4）， 265—266
8	Sun J.， Modern Chemical Research， 2017，（9）， 4—5
	孙静. 当代化工研究， 2017，（9）， 4—5
9	Mizuno H.， Ueda K.， Kobayashi Y.， Tsuyama N.， Todoroki K.， Min J. Z.， Toyo'oka T.， Biomed. Chromatogr.， 2017， 31（1）， e3864
10	Guan T.， Xin Y.， Zheng K.， Wang R. J.， Zhang X.， Jia S. Q.， Li S. Q.， Cao C.， Zhao X. J.， Biometals， 2021， 34（1）， 33—48
11	Tan B. B.， Zhang Y.， Zhang T. T.， He J. S.， Luo X. Y.， Bian X. Q.， Wu J. L.， Zou C.， Wang Y. Z.， Fu L.， Biomed. Chromatogr.， 2020， 34（10）， e4922
12	Yang B. Y.， Li X. L.， Yao Y.， Wang C.， Chen S. N.， Wu X. J.， Wu D. P.， J. Exp. Hematol.， 2017， 25（6）， 1585—1591
	杨冰玉，李晓莉，姚遥，王畅，陈苏宁，吴小津，吴德沛. 中国实验血液学杂志， 2017， 25（6）， 1585—1591
13	Liebal U. W.， Phan A. N. T.， Sudhakar M.， Raman K.， Blank L. M.， Metabolites， 2020， 10（6）， 243
14	Wang W. Y.， He Z. R.， Kong Y.， Liu Z. Q.， Gong L. Z.， Clin. Chim. Acta， 2021， 519， 10—17
15	Baiges⁃Gaya G.， Iftimie S.， Castañé H.， Rodríguez⁃Tomàs E.， Jiménez⁃Franco A.， López⁃Azcona A. F.， Castro A.， Camps J.， Joven J.， Biomolecules， 2023， 13（1）， 163
16	Jin Q. Q.， Sun J. H.， Du Q. X.， Lu X. J.， Zhu X. Y.， Fan H. L.， Hölscher C.， Wang Y. Y.， Int. J. Mol. Med.， 2017， 40（4）， 1019—1028
17	Fan S.， Kind T.， Cajka T.， Hazen S. L.， Tang W. H. W.， Kaddurah⁃Daouk R.， Irvin M. R.， Arnett D. K.， Barupal D. K.， Fiehn O.， Anal. Chem.， 2019， 91（5）， 3590—3596
18	Cao J.， Lv X. G.， Li Y.， Jin Q. Q.， Chu X. Y.， Wang Y. Y.， Sun J. H.， J. Forensic Med.， 2018， 34（3）， 228—232
	曹洁，吕晓革，李宇，靳茜茜，储晓云，王英元，孙俊红. 法医学杂志， 2018， 34（3）， 228—232
19	Pan H.， Exploring the Formation Mechanism of Deep Vein Thrombosis of Lower Extremities Based on Multi⁃omics Technology， Jilin University， Changchun， 2021
	潘赫. 基于多组学技术探讨下肢深静脉血栓的形成机制，长春：吉林大学， 2021
20	Zhou Y. S.， Qu J. L.， Han Y. L.， Chin. J. Cell Biol.， 2023， 45（3）， 428—435
	周悦思，曲家良，韩英伦. 中国细胞生物学学报， 2023， 45（3）， 428—435
21	Branchford B. R.， Carpenter S. L.， Front. Pediatr.， 2018， 6， 142
22	Kursa M. B.， Rudnicki W. R.， J. Stat. Softw.， 2010， 36（11）， 1—13
23	Nian Y.， Su X. X.， Yue H.， Zhu Y. J.， Li J.， Wang W. Q.， Sheng Y. L.， Ma Q.， Liu J. K.， Li X. W.， Front. Plant Sci.， 2024， 15， 1396183

No.	Compound	Molecular formula	m/z	Retention time/min	Similarity	VIP	P value
1	2⁃Aminoisobutyric acid	C₄H₉NO₂	130.02	11.671	0.957	1.595	0.003
2	Citric acid	C₆H₈O₇	272.98	29.074	0.935	1.731	<0.001
3	Nicotinamide	C₆H₆N₂O	178.97	19.154	0.935	1.641	<0.001
4	3⁃Methyl⁃2⁃oxovaleric acid	C₆H₁₀O₃	89.04	11.802	0.935	1.396	0.010
5	3⁃Deoxytetronic acid	C₄H₈O₄	117.02	15.240	0.957	1.425	0.003
6	3⁃Hydroxypropanoate	C₃H₆O₃	146.99	11.077	0.957	1.474	0.008
7	5⁃Oxoproline	C₅H₇NO₃	156.00	20.518	0.935	1.328	0.028
8	D⁃Pinitol	C₇H₁₄O₆	260.02	29.327	0.761	1.772	<0.001
9	L⁃Isoleucine	C₆H₁₃NO₂	158.00	13.780	0.891	1.500	0.003
10	L⁃Ornithine hydrochloride	C₅H₁₃ClN₂O₂	185.95	26.954	0.717	1.182	0.036
11	L⁃Proline	C₅H₉NO₂	142.00	14.289	0.935	1.294	0.021
12	N⁃Acetyl⁃L⁃glutamate	C₇H₁₁NO₅	185.99	19.805	0.935	1.404	0.011
13	β⁃D⁃Glucose	C₆H₁₂O₆	319.00	31.510	0.935	1.421	0.011
14	Glycolate	C₂H₄O₃	147.00	9.516	0.935	1.341	0.024
15	Decanoic acid	C₁₀H₂₀O₂	229.02	18.563	0.935	1.535	0.002
No.	Compound	Molecular formula	m/z	Retention time/min	Similarity	VIP	P value
16	Glycerophosphoric acid	C₃H₉O₆P	356.97	27.377	0.935	1.448	0.002
17	Isocitric acid	C₆H₈O₇	245.03	29.212	0.913	1.602	0.001
18	Suberic acid	C₈H₁₄O₄	117.98	25.755	0.935	1.645	<0.001
19	Allose	C₆H₁₂O₆	319.00	31.507	0.761	1.506	0.004
20	4⁃Chloropyridine⁃2，6⁃dicarboxylic Acid	C₇H₄ClNO₄	258.97	38.004	0.935	1.219	0.026
21	Conduritol⁃β⁃expoxide	C₆H₁₀O₅	317.96	34.204	0.935	1.425	0.005
22	Oxalacetic acid	C₄H₄O₅	174.00	8.887	0.935	1.206	0.042

No.	Compound	Molecular formula	m/z	Retention time/min	Similarity	VIP	P value
1	2⁃Aminoisobutyric acid	C₄H₉NO₂	130.02	11.671	0.957	1.595	0.003
2	Citric acid	C₆H₈O₇	272.98	29.074	0.935	1.731	<0.001
3	Nicotinamide	C₆H₆N₂O	178.97	19.154	0.935	1.641	<0.001
4	3⁃Methyl⁃2⁃oxovaleric acid	C₆H₁₀O₃	89.04	11.802	0.935	1.396	0.010
5	3⁃Deoxytetronic acid	C₄H₈O₄	117.02	15.240	0.957	1.425	0.003
6	3⁃Hydroxypropanoate	C₃H₆O₃	146.99	11.077	0.957	1.474	0.008
7	5⁃Oxoproline	C₅H₇NO₃	156.00	20.518	0.935	1.328	0.028
8	D⁃Pinitol	C₇H₁₄O₆	260.02	29.327	0.761	1.772	<0.001
9	L⁃Isoleucine	C₆H₁₃NO₂	158.00	13.780	0.891	1.500	0.003
10	L⁃Ornithine hydrochloride	C₅H₁₃ClN₂O₂	185.95	26.954	0.717	1.182	0.036
11	L⁃Proline	C₅H₉NO₂	142.00	14.289	0.935	1.294	0.021
12	N⁃Acetyl⁃L⁃glutamate	C₇H₁₁NO₅	185.99	19.805	0.935	1.404	0.011
13	β⁃D⁃Glucose	C₆H₁₂O₆	319.00	31.510	0.935	1.421	0.011
14	Glycolate	C₂H₄O₃	147.00	9.516	0.935	1.341	0.024
15	Decanoic acid	C₁₀H₂₀O₂	229.02	18.563	0.935	1.535	0.002
No.	Compound	Molecular formula	m/z	Retention time/min	Similarity	VIP	P value
16	Glycerophosphoric acid	C₃H₉O₆P	356.97	27.377	0.935	1.448	0.002
17	Isocitric acid	C₆H₈O₇	245.03	29.212	0.913	1.602	0.001
18	Suberic acid	C₈H₁₄O₄	117.98	25.755	0.935	1.645	<0.001
19	Allose	C₆H₁₂O₆	319.00	31.507	0.761	1.506	0.004
20	4⁃Chloropyridine⁃2，6⁃dicarboxylic Acid	C₇H₄ClNO₄	258.97	38.004	0.935	1.219	0.026
21	Conduritol⁃β⁃expoxide	C₆H₁₀O₅	317.96	34.204	0.935	1.425	0.005
22	Oxalacetic acid	C₄H₄O₅	174.00	8.887	0.935	1.206	0.042

Pathway name	Total	Hits	P value	Impact
Glyoxylate and dicarboxylate metabolism	32	4	0.00013	0.13493
Citrate cycle （TCA cycle）	20	3	0.000612	0.25318
Alanine， aspartate and glutamate metabolism	28	2	0.025050	0.09135

Pathway name	Total	Hits	P value	Impact
Glyoxylate and dicarboxylate metabolism	32	4	0.00013	0.13493
Citrate cycle （TCA cycle）	20	3	0.000612	0.25318
Alanine， aspartate and glutamate metabolism	28	2	0.025050	0.09135

[1]	ZHANG Yan, JIANG Xingjian, LIU Ming, ZHENG Zhi, ZHANG Yong. Predict Efficiency of Organic Solar Cell with Low Generalization Error Based on Molecular Property and Device Fabrication [J]. Chem. J. Chinese Universities, 2023, 44(7): 20230165.
[2]	WONG Honho, LU Qiuyang, SUN Mingzi, HUANG Bolong. Rational Design of Graphdiyne-based Atomic Electrocatalysts： DFT and Self-validated Machine Learning [J]. Chem. J. Chinese Universities, 2022, 43(5): 20220042.
[3]	DU Jiahui, LIAO Kang, HONG Benkun, WANG Zhongye, MA Jing, LI Wei, LI Shuhua. Generalized Energy-based Fragmentation Clustering Algorithm for Localized Excited States [J]. Chem. J. Chinese Universities, 2021, 42(7): 2227.
[4]	LI Wenwen, ZHU Airu, LONG Yijing, WANG Chunyan, HAN Yuanping, DUAN Yixiang. Metabolomics Study of Serum and Liver in Type 2 Diabetes Mice Induced by High Fat Diet with Vitamin D Deficiency^† [J]. Chem. J. Chinese Universities, 2018, 39(11): 2395.
[5]	HUANG Yu, GU Caiyun, WU Hanzhong, XIA Xiaoshuang, LI Xin. UPLC-QTOF-MS-based Metabolomics Study on Ischemic Stroke Patients^† [J]. Chem. J. Chinese Universities, 2017, 38(10): 1742.
[6]	LI Penghui, DENG Lingli, LUO Jiao, LI Wei, NING Jing, DING Jianhua, WU Xiaoping. EESI-MS Detection and Statistical Analysis of Multi-batch of Exhaled Breath Metabolomics Data of Liver Failure Patients [J]. Chem. J. Chinese Universities, 2016, 37(4): 626.
[7]	ZHANG Ruixing, LIU Shu, PI Zifeng, SONG Fengrui, LIU Zhiqiang. Cell Metabolomics Study of Hg(Ⅱ) Ion Effect on the Metabolic Pathways of Cells [J]. Chem. J. Chinese Universities, 2014, 35(6): 1146.
[8]	HUANG Min, GAO Jian-Yi, ZHAI Zhi-Gang, LIANG Qiong-Lin, WANG Yi-Ming, BAI Yan-Qiang, LUO Guo-An. Influence of Caffeine on the Sedative Effects of Promethazine Based on Pharmacokinetics and Metabolomics [J]. Chem. J. Chinese Universities, 2013, 34(4): 829.
[9]	MEI Zhen, CAI Wen-Sheng, SHAO Xue-Guang. Rapid Analysis of Complex Samples by GC-MS with Immune Algorithm [J]. Chem. J. Chinese Universities, 2013, 34(4): 819.
[10]	WU Sheng-Ming, FENG Bo, CHENG Jian-Hua, LI Hai-Jing, FANG Jun-Jian, YAN Xian-Zhong, WEI Lai, DONG Fang-Ting. GC/TOF-MS-based Analytical Method of Endogenous Metabolites for Mouse Serum [J]. Chem. J. Chinese Universities, 2012, 33(06): 1188.
[11]	DONG Ji-Yang, LI Wei, DENG Ling-Li, XU Jing-Jing, Julian L. Griffin, CHEN Zhong*. New Variable Scaling Method for NMR-based Metabolomics Data Analysis [J]. Chem. J. Chinese Universities, 2011, 32(2): 262.
[12]	SUN Shuo, ZHAI Yu-Juan, SUN Ye, ZHANG Yu-Pu, LIU He, WANG Xing-Hua, YU Ai-Min, . Application of Ionic Liquid-non Polar Solvent Microwave Extraction to the Study of the Chemical Constituents from Roots of Panax ginseng* C. M. Mey. [J]. Chem. J. Chinese Universities, 2010, 31(3): 468.
[13]	HUANG Xin, GONG Yi-Fei, WANG Yi*, QU Hai-Bin, CHENG Yi-Yu. Metabonomic Study of Hepato-protective Effect of Silybin for the Carbon Tetrachloride Induced Liver Injure in Mice [J]. Chem. J. Chinese Universities, 2008, 29(4): 714.
[14]	YI Lun-Zhao, LIANG Yi-Zeng, ZENG Zhong-Da, WANG Ping, YUAN Da-Lin. Comparative Analysis of Volatile Constituents in Chenpi of Different Original Plants by GC-MS and AMWFA [J]. Chem. J. Chinese Universities, 2006, 27(9): 1626.
[15]	LI Xiao-Ru, LIANG Yi-Ceng, YANG Hui, GUO Fang-Qiu, LI Xiao-Ning, ZENG Zhong-Da. Chemical Component Study of Herb Pairs in Traditional Chinese Medicine --Analysis of Volatile Oil in Rhizoma Ligustici Chuanxiong-Radix Paeoniae Rubra by GC-MS and Chemometric Resolution Method [J]. Chem. J. Chinese Universities, 2006, 27(3): 443.