Table 3.

The 23 identified biomarkers of NAFLD mice treated by DWYG and its single herb with their related information.

No.	Metabolitesa	VIPb	Trendc							Metabolic pathwaysd
			M/C	AC/M	CL/M	SC/M	RG/M	GC/M	DW/M
1	Valerylcarnitine	1.230	↑	–	–	–	–	↓	↓	Fatty acid beta-oxidation
2	N-Acetyldopamine	1.365	↑	↓	↓	↓	↓	↓	↓	Amino acid metabolism
3	N-Palmitoyl Arginine	1.309	↑**	↓	–	–	–	–	↓	Amino acid metabolism
4	Sphingosine	1.361	↑*	–	–	↓	↓	–	↓	Sphingolipid signaling pathway
5	Arachidonoylcarnitine	1.178	↑**	↓	–	–	–	–	–	Fatty acid beta-oxidation
6	6-Methylene-4-pregnene-3,20-dione	1.099	↓	↑	↑	–	↑	↑	–	Steroid hormone biosynthesis
7	LysoPA (22:6)	1.050	↓	–	↑**	↑**	↑**	↑	↑**	Glycerophospholipid metabolism
8	LysoPC (22:5)	1.578	↓**	–	↑*	↑**	↑**	↑**	↑**	Glycerophospholipid metabolism
9	LysoPC (20:3)	1.649	↑**	↓	–	↓**	–	↓**	–	Glycerophospholipid metabolism
10	PG (TXB2/i-24:0)	1.818	↓*	–	↑	↑	–	–	↑	Glycerophospholipid metabolism
11	LysoPE (16:0)	1.741	↓**	–	↑**	↑**	↑*	–	↑**	Glycerophospholipid metabolism
12	MG (22:6/0:0/0:0)	1.869	↓**	–	↑	↑**	↑*	–	↑**	Glycerophospholipid metabolism
13	MG (18:2/0:0/0:0)	2.073	↓**	–	↑	↑**	↑*	–	↑**	Glycerophospholipid metabolism
14	PC (20:5-3OH/2:0)	1.881	↑**	↓**	↓**	↓**	–	↓**	↓*	Glycerophospholipid metabolism
15	LysoPC (20:2)	1.690	↑**	↓**	↓**	↓**	–	↓**	↓**	Glycerophospholipid metabolism
16	5beta-Pregnane-3,20-dione	1.111	↓**	–	–	–	↑	–	–	Steroid hormone biosynthesis
17	PA (20:4-2OH/i-14:0)	1.508	↓**	↓**	↓**	↓**	↓**	↓**	↓**	Glycerophospholipid metabolism
18	LysoPE (18:0)	1.748	↓	–	↑	↑	–	–	↑	Glycerophospholipid metabolism
19	LysoPI (20:4)	1.395	↑**	–	↓**	↓	–	–	↓	Glycerophospholipid metabolism
20	Calcium octadecanoate	1.074	↑*	–	↓**	↓	–	–	↓*	Biosynthesis of unsaturated fatty acids
21	Ethyl octadec-9-enoate	1.688	↓*	–	–	↑	–	–	↑	Biosynthesis of unsaturated fatty acids
22	dihydrocorticosterone	1.302	↓	–	↑	↑	↑*	–	↑	Steroid hormone biosynthesis
23	Cerd (18:0/18:0)	1.256	↑**	–	↓**	↓*	–	–	↓**	Sphingolipid signaling pathway

*P < 0.05, **P < 0.01, significantly different from model group.

^aMetabolite identification was carried out by comparison to the METLIN and HMDB Metabolite Databases.

^bCalculated using the OPLS-DA model based on metabolites of control group and model group of serum samples; Selection was based on VIP values (threshold >1).

^cUp arrows indicate metabolites were upregulated, and down arrows indicate metabolites were downregulated; Selected metabolites based on FC ≥ 1.1 or FC ≤ 0.9.

^dRelated pathway was referred to KEGG.