Leveraging metabolic similarity in a ¹H NMR database of medicinal plants to advance pharmacognostic insights

Abstract

Natural products remain a central resource for drug discovery, and increasing evidence suggests that therapeutic effects often arise from the combined action of multiple constituents rather than single compounds. In this context, metabolomic profiling is essential for comparing complex plant chemical phenotypes, and ¹H NMR provides robust whole-profile fingerprints that support cross-species metabolic barcoding and systematic comparison. In this study, we establish and apply a standardized large-scale ¹H NMR database to enable macroscopic metabolomic similarity profiling of medicinal plants. Specifically, using ¹H NMR profiles from 656 traditional medicinal herbs, we demonstrate how this standardized large-scale metabolomic framework can be applied to key challenges in medicinal plant research, including quality control across different locations and time periods, identification of metabolically similar alternative species, and compositional analysis of multi-herb formulations. Our findings demonstrate the utility of this NMR-based strategy as a scalable approach for standardization, authentication, and holistic characterization of medicinal plants, advancing the field beyond reductionist paradigms. This study establishes a standardized large-scale ¹H NMR database of medicinal plants and introduces a macroscopic framework for large-scale metabolomic similarity profiling that enables chemotaxonomic contextualization, quality surveillance, and identification of metabolically similar candidate substitutes.

Supplementary Information

The online version contains supplementary material available at 10.1038/s41598-026-37725-2.

Introduction

Natural products have long been vital for human health, providing amongst other things, herbal medicines with centuries of empirical knowledge as a proof of their efficacy. With the advancement of modern science, it has become essential to scientifically explain and validate these traditional remedies. The discovery of active compounds such as strychnine¹, emetine², quinine³, and various terpenoids⁴ has provided a scientific foundation for the efficacy of these natural products. Notable examples of natural product-based medicines include aspirin, derived from salicylic acid in willow bark^5,6, and artemisinin, an antimalarial drug extracted from Artemisia annua⁷. These cases underscore the significant role that natural bioactive substances play in pharmaceutical innovation. As research progresses, natural products will continue to be a crucial resource for the development of new drugs and treating diseases through their diverse biological actions.

The diversity of chemical compounds in natural products is a unique opportunity for drug discovery and development. However, paradoxically, this chemical diversity has proved to be the major challenge to the possibility of taking full advantage of its potential as the identification and isolation of specific bioactive compounds from among the numerous chemical substances present in natural products, is often considerably time and resource consuming. Thus, in order to tackle this issue, natural product research demands the use of sophisticated analytical techniques and methods to evaluate their bioactivity, requiring a multifaceted approach.

Chemical profiling for these active compounds is currently a critical step in modern drug development, highlighting the significant role that natural bioactive substances can play in pharmaceutical innovation. Metabolomics and untargeted chemical profiling methods allow for the detection of numerous metabolites in biological systems without the time-consuming steps of compound isolation, facilitating both the qualitative and quantitative analysis of bioactive compounds. In this study, we establish and apply a standardized large-scale ¹H NMR database to enable macroscopic metabolomic similarity profiling of medicinal plants, providing an integrated framework for chemotaxonomic contextualization, quality surveillance, and the evaluation of multi-herb formulations. Throughout this study, metabolomic fingerprinting refers to whole spectrum of ¹H NMR profiles used as reproducible chemical fingerprints, while metabolic barcoding denotes their use for authentication and identification in a chemotaxonomic context. We use the term macroscopic to describe database-driven comparisons of overall chemical relatedness across many samples through large-scale metabolomic similarity profiling.

Analytical platforms used in chemical profiling must possess key features: reproducibility, high sensitivity, and high resolution. Reproducibility is critical for robust fingerprinting, as it ensures consistent results under identical experimental conditions - an essential factor for the reliability of research findings. High sensitivity and resolution are necessary for the accurate identification of target compounds, enabling precise structural characterization and the detection of subtle differences among analytes. Among the various analytical techniques available, mass spectrometry (MS)^8,9 and nuclear magnetic resonance (NMR)^10,11 have received considerable attention for meeting these requirements. Each method has its own advantages and limitations. Notably, NMR is a non-destructive technique that provides highly reproducible spectra, allowing for consistent results across replicates and even from long-term stored samples. Furthermore, NMR gives unique spectra that help establish the metabolic fingerprints of plants to characterise them. This long-term reproducibility of NMR-generated data results in NMR-based metabolomics databases that are frontrunners in terms of relevance if compared to those based on other analytical techniques. Along with its suitability for databases, NMR-based metabolomics has been successfully employed in the discovery of various bioactive substances due to its powerful identification capabilities^12,13. The number of published studies using NMR-based metabolomics attest to its standing as a highly efficient tool for the discovery of new drugs from natural products, highlighting its potential in the systematic analysis of complex chemical compositions and the identification of promising drug candidates^14–16.

Recently, there has been a shift in the approach for therapeutical solutions, from single drugs to complex mixtures of compounds. Given the growing importance of natural products in drug development, the complex interactions between the multiple components of natural products cannot be ignored when searching for the mechanisms behind their activity¹⁷. A possible explanation for the activity of these complex mixtures versus isolated compounds sourced from natural products, may be that they are involved in synergistic effects, providing more effective alternatives for the treatment of various diseases. This underscores the close connection between natural product research and complex drug development, indicating that natural products will continue to be a vital source for new pharmaceutical discoveries.

In agreement with this idea, the efficacy of traditional herbal medicines has not been typically attributed to a single component or a common set of components but rather to the interactions among an entire array of active constituents^18,19. This, which is in itself complex, is further complicated when considering that these metabolites are produced through unique biosynthetic pathways specific to each species, which can result in variations in component profiles and compositions based on origin, cultivation conditions, and other factors. Additionally, the therapeutic effects of one same medicinal herb can vary depending on factors such as the formulation, dosage, and preparation methods. This complexity has historically led to a reliance on empirical evidence for assessing the efficacy of herbal medicines, an approach which does not provide the necessary information on which to base their standardization from a modern pharmacological perspective. In this context, revisiting NMR profiling holds significant value for herbal medicine research as it has proved to be a powerful tool for the precise analysis of complex herbal components, helping to understand interactions among all constituents and discover new pharmacologically active substances²⁰. This approach can play a crucial role in advancing the study and use of herbal medicines.

The foundation of this investigation was to identify metabolic similarities among different plant species. NMR analysis proved particularly well-suited for this purpose, as it allows for the detection of functional group-specific signals across a broad spectrum of molecules, thereby facilitating the recognition of shared metabolic features among species. Importantly, NMR enables the quantitative comparison of metabolite profiles, offering a promising approach for exploring similarities in traditional herbal medicines. Many traditional herbs have well-documented therapeutic efficacy, which often arises from the synergistic action of multiple compounds rather than a single active metabolite.

Despite the unique advantages of NMR, its application in metabolomics-based natural products research has often been underestimated in previous studies. It has frequently been treated as interchangeable with other analytical techniques, particularly mass spectrometry (MS)-based approaches, without full recognition of its distinct strengths. To address this gap, we conducted a systematic comparison of metabolic similarities among over 656 established traditional herbal medicines (Table 1) using NMR analysis. This approach aims to identify shared chemical components and investigate their potential associations with specific therapeutic effects. The detected similarities were further explored for their implications in quality control, domestic substitution, and the chemical characterization of multi-herbal formulations. By elucidating the metabolic similarities and differences among these herbs, our findings contribute to a deeper understanding of how specific combinations of compounds in complex herbal mixtures may enhance therapeutic efficacy. Ultimately, this research seeks to reassess the value of traditional medicines through a modern scientific lens and to support the development of new drugs based on natural products.

Table 1.

Medicinal plant samples analyzed by ¹H NMR, including the information of sample codes, scientific names (family, genus, and species), and used parts.

No.	Sample code	Family	Genus	Species	Used part
1	NPL-KHU-001	Fabaceae	Pueraria	lobata	Root
2	NPL-KHU-002	Zingiberaceae	Curcuma	longa	Root
3	NPL-KHU-003	Zingiberaceae	Zingiber	officinale	Root
4	NPL-KHU-004	Anacardiaceae	Rhus	verniciflua	Rind
5	NPL-KHU-005	Nymphaeaceae	Euryale	ferox	Seed
6	NPL-KHU-006	Lauraceae	Cinnamomum	cassia	Rind
7	NPL-KHU-007	Lauraceae	Cinnamomum	cassia	Rind
8	NPL-KHU-008	Zingiberaceae	Alpinia	officinarum	Root
9	NPL-KHU-009	Fabaceae	Sophora	flavescens	Root
10	NPL-KHU-010	Viscaceae	Viscum	coloratum	Whole plant
11	NPL-KHU-011	Poaceae	Oryza	sativa	Seed
12	NPL-KHU-012	Laminariaceae	Laminaria	japonica	Whole plant
13	NPL-KHU-013	Lamiaceae	Agastache	rugosa	Whole plant
14	NPL-KHU-014	Fabaceae	Sophora	japonica	Flower and seed
15	NPL-KHU-015	Celastraceae	Euonymus	alatus	Rind
16	NPL-KHU-016	Rosaceae	Rosa	laevigata	Fruit
17	NPL-KHU-017	Balsaminaceae	Impatiens	balsamina	Whole plant
18	NPL-KHU-018	Poaceae	Oryza	sativa	Seed
19	NPL-KHU-019	Fabaceae	Vigna	radiatus	Seed
20	NPL-KHU-020	Lamiaceae	Salvia	miltiorrhiza	Root
21	NPL-KHU-021	Poaceae	Lophatherum	gracile	Whole plant
22	NPL-KHU-022	Apiaceae	Angelica	gigas	Root
23	NPL-KHU-023	Polygonaceae	Rheum	palmatum	Root
24	NPL-KHU-024	Cucurbitaceae	Benincasa	cerifera	Fruit
25	NPL-KHU-025	Fabaceae	Glycine	max	Seed
26	NPL-KHU-026	Rosaceae	Rosa	rugosa	Fruit
27	NPL-KHU-027	Malvaceae	Hibiscus	syriacus	Flower
28	NPL-KHU-028	Paeoniaceae	Paeonia	suffruticosa	Root
29	NPL-KHU-029	Cucurbitaceae	Momordica	cochinchinensis	Fruit
30	NPL-KHU-030	Actinidiaceae	Actinidia	polygama	Fruit
31	NPL-KHU-031	Lamiaceae	Mentha	arvensis	Leaf
32	NPL-KHU-032	Lamiaceae	Scutellaria	barbata	Whole plant
33	NPL-KHU-033	Araceae	Pinellia	ternata	Root
34	NPL-KHU-034	Menispermaceae	Sinomenium	acutum	Root
35	NPL-KHU-035	Ginkgoaceae	Ginkgo	biloba	Leaf
36	NPL-KHU-036	Cupressaceae	Platycladus	orientalis	Seed and leaf
37	NPL-KHU-037	Paeoniaceae	Paeonia	japonica	Root
38	NPL-KHU-038	Rubiaceae	Oldenlandia	herbacea	Whole plant
39	NPL-KHU-039	Poaceae	Triticum	aestivum	Seed
40	NPL-KHU-040	Lemnaceae	Spirodela	polyrhiza	Whole plant
41	NPL-KHU-041	Dioscoreaceae	Dioscorea	tokoro	Root
42	NPL-KHU-042	Arecaceae	Areca	catechu	Seed
43	NPL-KHU-043	Cucurbitaceae	Cucumis	melo	Fruit
44	NPL-KHU-044	Combretaceae	Quisqualis	indica	Flower
45	NPL-KHU-045	Apiaceae	Aegopodium	podagraria	Whole plant
46	NPL-KHU-046	Cornaceae	Cornus	officinalis	Fruit
47	NPL-KHU-047	Lythraceae	Punica	granatum	Fruit
48	NPL-KHU-048	Zingiberaceae	Elettaria	cardamomum	Seed
49	NPL-KHU-049	Apiaceae	Anethum	graveolens	Whole plant
50	NPL-KHU-050	Ebenaceae	Diospyros	kaki	Fruit
51	NPL-KHU-051	Apiaceae	Bupleurum	falcatum	Root
52	NPL-KHU-052	Polygonaceae	Rumex	japonicus	Whole plant
53	NPL-KHU-053	Primulaceae	Lysimachia	foenum-graecum	Whole plant
54	NPL-KHU-054	Lardizabalaceae	Akebia	quinata	Fruit and stem
55	NPL-KHU-055	Crassulaceae	Orostachys	japonica	Whole plant
56	NPL-KHU-056	Caryophyllaceae	Silene	firma	Whole plant
57	NPL-KHU-057	Solanaceae	Solanum	nigrum	Whole plant
58	NPL-KHU-058	Asteraceae	Arctium	lappa	Root
59	NPL-KHU-059	Brassicaceae	Brassica	napus	Seed
60	NPL-KHU-060	Asteraceae	Artemisia	anomala	Leaf
61	NPL-KHU-061	Ulmaceae	Ulmus	macrocarpa	Rind
62	NPL-KHU-062	Burseraceae	Boswellia	carteri	Resin
63	NPL-KHU-063	Lauraceae	Cinnamomum	cassia	Rind
64	NPL-KHU-064	Orobanchaceae	Cistanche	deserticola	Whole plant
65	NPL-KHU-065	Asteraceae	Artemisia	capillaris	Leaf
66	NPL-KHU-066	Apiaceae	Angelica	acutiloba	Root
67	NPL-KHU-067	Fabaceae	Pterocarpus	santalinus	Wood
68	NPL-KHU-068	Lamiaceae	Perilla	frutescens	Leaf
69	NPL-KHU-069	Lauraceae	Cinnamomum	camphora	Leaf and rind
70	NPL-KHU-070	Polyporaceae	Polyporus	umbellatus	Mushroom
71	NPL-KHU-071	Urticaceae	Boehmeria	nivea	Stem
72	NPL-KHU-072	Adoxaceae	Sambucus	williamsii	Fruit
73	NPL-KHU-073	Tamaricaceae	Tamarix	chinensis	Whole plant
74	NPL-KHU-074	Myrtaceae	Syzygium	aromaticum	Flower bud
75	NPL-KHU-075	Campanulaceae	Adenophora	remotiflora	Root
76	NPL-KHU-076	Rhamnaceae	Hovenia	dulcis	Fruit
77	NPL-KHU-077	Amaranthaceae	Kochia	scoparia	Whole plant
78	NPL-KHU-078	Rutaceae	Poncirus	trifoliata	Fruit
79	NPL-KHU-079	Plantaginaceae	Plantago	asiatica	Seed and leaf
80	NPL-KHU-080	Apiaceae	Cnidium	officinale	Root
81	NPL-KHU-081	Apiaceae	Ligusticum	chuanxiong	Root
82	NPL-KHU-082	Amaranthaceae	Celosia	argentea	Whole plant
83	NPL-KHU-083	Malvaceae	Alcea	rosea	Flower
84	NPL-KHU-084	Amaryllidaceae	Allium	fistulosum	Whole plant
85	NPL-KHU-085	Lamiaceae	Leonurus	japonicus	Whole plant
86	NPL-KHU-086	Cupressaceae	Platycladus	orientalis	Rind
87	NPL-KHU-087	Rubiaceae	Gardenia	jasminoides	Fruit
88	NPL-KHU-088	Thymelaeaceae	Aquilaria	agallocha	Resin
89	NPL-KHU-089	Apiaceae	Carum	carvi	Seed
90	NPL-KHU-090	Apiaceae	Cuminum	cyminum	Seed
91	NPL-KHU-091	Schisandraceae	Illicium	verum	Seed
92	NPL-KHU-092	Euphorbiaceae	Ricinus	communis	Seed
93	NPL-KHU-093	Nelumbonaceae	Nelumbo	nucifera	Flower, seed and root
94	NPL-KHU-094	Asteraceae	Carpesium	abrotanoides	Whole plant
95	NPL-KHU-095	Asteraceae	Eclipta	prostrata	Whole plant
96	NPL-KHU-096	Lamiaceae	Phlomis	umbrosa	Whole plant
97	NPL-KHU-097	Asteraceae	Artemisia	gmelinii	Leaf
98	NPL-KHU-098	Lygodiaceae	Lygodium	japonicum	Whole plant
99	NPL-KHU-099	Apiaceae	Glehnia	littoralis	Root
100	NPL-KHU-100	Pinaceae	Pinus	koraiensis	Seed
101	NPL-KHU-101	Lamiaceae	Elsholtzia	ciliata	Whole plant
102	NPL-KHU-102	Geraniaceae	Geranium	thunbergii	Root
103	NPL-KHU-103	Lamiaceae	Nepeta	tenuifolia	Whole plant
104	NPL-KHU-104	Juglandaceae	Juglans	regia	Seed and rind
105	NPL-KHU-105	Asteraceae	Carthamus	tinctorius	Flower
106	NPL-KHU-106	Lamiaceae	Scutellaria	baicalensis	Root
107	NPL-KHU-107	Lauraceae	Lindera	obtusiloba	Whole plant
108	NPL-KHU-108	Apiaceae	Foeniculum	vulgare	Seed
109	NPL-KHU-109	Fabaceae	Glycine	max	Seed
110	NPL-KHU-110	Pedaliaceae	Sesamum	indicum	Seed
111	NPL-KIOM-001	Fabaceae	Pueraria	lobata	Root
112	NPL-KIOM-002	Asteraceae	Chrysanthemum	indicum	Flower
113	NPL-KIOM-003	Fabaceae	Glycyrrhiza	glabra	Root
114	NPL-KIOM-004	Fabaceae	Glycyrrhiza	inflata	Root
115	NPL-KIOM-005	Fabaceae	Glycyrrhiza	uralensis	Root
116	NPL-KIOM-006	Apiaceae	Notopterygium	forbesii	Root
117	NPL-KIOM-007	Apiaceae	Notopterygium	incisum	Root
118	NPL-KIOM-008	Apiaceae	Ostericum	koreanum	Root
119	NPL-KIOM-009	Zingiberaceae	Curcuma	longa	Root
120	NPL-KIOM-010	Brassicaceae	Brassica	juncea	Seed and leaf
121	NPL-KIOM-011	Zingiberaceae	Zingiber	officinale	Root
122	NPL-KIOM-012	Fagaceae	Castanea	crenata	Fruit
123	NPL-KIOM-013	Nymphaeaceae	Euryale	ferox	Seed
124	NPL-KIOM-014	Fabaceae	Cassia	obtusifolia	Seed
125	NPL-KIOM-015	Fabaceae	Cassia	tora	Seed
126	NPL-KIOM-016	Lauraceae	Cinnamomum	cassia	Rind
127	NPL-KIOM-017	Zingiberaceae	Alpinia	officinarum	Root
128	NPL-KIOM-018	Apiaceae	Ligusticum	sinense	Root
129	NPL-KIOM-019	Apiaceae	Ligusticum	tenuissima	Root
130	NPL-KIOM-020	Santalaceae	Viscum	album	Whole plant
131	NPL-KIOM-021	Fabaceae	Caragana	sinica	Root
132	NPL-KIOM-022	Polypodiaceae	Drynaria	fortunei	Root
133	NPL-KIOM-023	Lamiaceae	Agastache	rugosa	Whole plant
134	NPL-KIOM-024	Cucurbitaceae	Trichosanthes	kirilowii	Root
135	NPL-KIOM-025	Cucurbitaceae	Trichosanthes	rosthornii	Root
136	NPL-KIOM-026	Cucurbitaceae	Trichosanthes	kirilowii	Root
137	NPL-KIOM-027	Cucurbitaceae	Trichosanthes	rosthornii	Root
138	NPL-KIOM-028	Lamiaceae	Pogostemon	cablin	Whole plant
139	NPL-KIOM-029	Solanaceae	Lycium	barbarum	Fruit
140	NPL-KIOM-030	Solanaceae	Lycium	chinense	Fruit
141	NPL-KIOM-031	Asteraceae	Chrysanthemum	zawadskii	Flower
142	NPL-KIOM-032	Asteraceae	Chrysanthemum	zawadskii	Flower
143	NPL-KIOM-033	Dicksoniaceae	Cibotium	barometz	Root
144	NPL-KIOM-034	Asteraceae	Chrysanthemum	morifolium	Flower
145	NPL-KIOM-035	Caprifoliaceae	Lonicera	japonica	Flower and leaf
146	NPL-KIOM-036	Primulaceae	Lysimachia	christinae	Root
147	NPL-KIOM-037	Campanulaceae	Platycodon	grandiflorum	Root
148	NPL-KIOM-038	Brassicaceae	Raphanus	sativus	Root
149	NPL-KIOM-039	Poaceae	Phragmites	communis	Whole plant
150	NPL-KIOM-040	Lamiaceae	Salvia	miltiorrhiza	Root
151	NPL-KIOM-041	Apiaceae	Angelica	gigas	Root
152	NPL-KIOM-042	Campanulaceae	Codonopsis	pilosula	Root
153	NPL-KIOM-043	Campanulaceae	Codonopsis	tangshen	Root
154	NPL-KIOM-044	Campanulaceae	Codonopsis	pilosula	Root
155	NPL-KIOM-045	Arecaceae	Areca	catechu	Seed
156	NPL-KIOM-046	Rhamnaceae	Zizyphus	jujuba	Fruit
157	NPL-KIOM-047	Rhamnaceae	Zizyphus	jujuba	Fruit
158	NPL-KIOM-048	Rosaceae	Prunus	davidiana	Seed
159	NPL-KIOM-049	Rosaceae	Prunus	persica	Fruit
160	NPL-KIOM-050	Araliaceae	Aralia	continentalis	Root
161	NPL-KIOM-051	Eucommiaceae	Eucommia	ulmoides	Rind
162	NPL-KIOM-052	Eucommiaceae	Eucommia	ulmoides	Rind
163	NPL-KIOM-053	Cannabaceae	Cannabis	sativa	Whole plant
164	NPL-KIOM-054	Portulacaceae	Portulaca	oleracea	Whole plant
165	NPL-KIOM-055	Ephedraceae	Ephedra	sinica	Whole plant
166	NPL-KIOM-056	Ephedraceae	Ephedra	intermedia	Whole plant
167	NPL-KIOM-057	Ephedraceae	Ephedra	equisetina	Whole plant
168	NPL-KIOM-058	Ephedraceae	Ephedra	intermedia	Whole plant
169	NPL-KIOM-059	Ephedraceae	Ephedra	equisetina	Whole plant
170	NPL-KIOM-060	Verbenaceae	Vitex	rotundifolia	Seed
171	NPL-KIOM-061	Verbenaceae	Vitex	trifolia	Seed
172	NPL-KIOM-062	Asparagaceae	Liriope	platyphylla	Root
173	NPL-KIOM-063	Asparagaceae	Ophiopogon	japonicus	Root
174	NPL-KIOM-064	Poaceae	Hordeum	vulgare	Whole plant
175	NPL-KIOM-065	Rosaceae	Chaenomeles	sinensis	Fruit
176	NPL-KIOM-066	Rosaceae	Chaenomeles	speciosa	Fruit
177	NPL-KIOM-067	Paeoniaceae	Paeonia	suffruticosa	Root
178	NPL-KIOM-068	Equisetaceae	Equisetum	hyemale	Whole plant
179	NPL-KIOM-069	Lardizabalaceae	Akebia	quinata	Fruit
180	NPL-KIOM-070	Asteraceae	Aucklandia	lappa	Root
181	NPL-KIOM-071	Lamiaceae	Mentha	arvensis	Whole plant
182	NPL-KIOM-072	Araceae	Pinellia	ternata	Root
183	NPL-KIOM-073	Menispermaceae	Sinomenium	acutum	Root
184	NPL-KIOM-074	Apiaceae	Saposhnikovia	divaricata	Root
185	NPL-KIOM-075	Ginkgoaceae	Ginkgo	biloba	Leaf
186	NPL-KIOM-076	Papaveraceae	Chelidonium	majus	Whole plant
187	NPL-KIOM-077	Zingiberaceae	Amomum	kravanh	Seed
188	NPL-KIOM-078	Zingiberaceae	Amomum	compactum	Seed
189	NPL-KIOM-079	Apocynaceae	Cynanchum	wilfordii	Root
190	NPL-KIOM-080	Apiaceae	Angelica	dahurica	Root
191	NPL-KIOM-081	Apiaceae	Angelica	dahurica	Root
192	NPL-KIOM-082	Asteraceae	Atractylodes	japonica	Root
193	NPL-KIOM-083	Asteraceae	Atractylodes	macrocephala	Root
194	NPL-KIOM-084	Fabaceae	Dolichos	lablab	Fruit
195	NPL-KIOM-085	Rubiaceae	Hedyotis	diffusa	Whole plant
196	NPL-KIOM-086	Fabaceae	Psoralea	corylifolia	Seed
197	NPL-KIOM-087	Polyporaceae	Poria	cocos	Mycelium
198	NPL-KIOM-088	Rosaceae	Rubus	coreanus	Fruit
199	NPL-KIOM-089	Ranunculaceae	Aconitum	carmichaeli	Root
200	NPL-KIOM-090	Ranunculaceae	Aconitum	carmichaeli	Root
201	NPL-KIOM-091	Arecaceae	Areca	catechu	Seed
202	NPL-KIOM-092	Campanulaceae	Adenophora	triphylla	Root
203	NPL-KIOM-093	Apiaceae	Cnidium	monieri	Seed
204	NPL-KIOM-094	Apiaceae	Torilis	japonica	Seed
205	NPL-KIOM-095	Zingiberaceae	Amomum	villosum	Seed
206	NPL-KIOM-096	Zingiberaceae	Amomum	villosum	Seed
207	NPL-KIOM-097	Rosaceae	Crataegus	pinnatifida	Fruit
208	NPL-KIOM-098	Cornaceae	Cornus	officinalis	Fruit
209	NPL-KIOM-099	Dioscoreaceae	Dioscorea	batatas	Root
210	NPL-KIOM-100	Dioscoreaceae	Dioscorea	japonica	Root
211	NPL-KIOM-101	Rhamnaceae	Zizyphus	jujuba	Fruit and root
212	NPL-KIOM-102	Rhamnaceae	Zizyphus	jujuba	Fruit and root
213	NPL-KIOM-103	Rutaceae	Zanthoxylum	piperitum	Rind and seed
214	NPL-KIOM-104	Rutaceae	Zanthoxylum	schinifolium	Rind and seed
215	NPL-KIOM-105	Rutaceae	Zanthoxylum	bungeanum	Rind and seed
216	NPL-KIOM-106	Sparganiaceae	Sparganium	stoloniferum	Root
217	NPL-KIOM-107	Moraceae	Morus	alba	Leaf and fruit
218	NPL-KIOM-108	Moraceae	Morus	alba	Leaf and fruit
219	NPL-KIOM-109	Moraceae	Morus	bombycis	Leaf and root
220	NPL-KIOM-110	Moraceae	Morus	alba	Leaf and fruit
221	NPL-KIOM-111	Orchidaceae	Dendrobium	fimbriatum	Whole plant
222	NPL-KIOM-112	Orchidaceae	Dendrobium	candidum	Whole plant
223	NPL-KIOM-113	Acoraceae	-	Acorus	gramineus
224	NPL-KIOM-114	Hypoxidaceae	Curculigo	orchioides	Root
225	NPL-KIOM-115	Aristolochiaceae	Asiasarum	heterotropoides	Root
226	NPL-KIOM-116	Aristolochiaceae	Asiasarum	sieboldii	Root
227	NPL-KIOM-117	Fabaceae	Cassia	angustifolia	Leaf and seed
228	NPL-KIOM-118	Fabaceae	Caesalpinia	sappan	Seed and rind
229	NPL-KIOM-119	Dipsacaceae	Dipsacus	asperoides	Root
230	NPL-KIOM-120	Ranunculaceae	Cimicifuga	heracleifolia	Root
231	NPL-KIOM-121	Ranunculaceae	Cimicifuga	dahurica	Root
232	NPL-KIOM-122	Apiaceae	Anethum	graveolens	Whole plant
233	NPL-KIOM-123	Ebenaceae	Diospyros	kaki	Fruit
234	NPL-KIOM-124	Apiaceae	Bupleurum	falcatum	Root
235	NPL-KIOM-125	Apiaceae	Bupleurum	scorzonerifolium	Root
236	NPL-KIOM-126	Apiaceae	Peucedanum	japonicum	Root
237	NPL-KIOM-127	Lycopodiaceae	Lycopodium	clavatum	Whole plant
238	NPL-KIOM-128	Zingiberaceae	Curcuma	kwangsiensis	Root
239	NPL-KIOM-129	Zingiberaceae	Curcuma	phaeocaulis	Root
240	NPL-KIOM-130	Zingiberaceae	Curcuma	wenyujin	Root
241	NPL-KIOM-131	Asteraceae	Artemisia	argyi	Leaf
242	NPL-KIOM-132	Asteraceae	Artemisia	princeps	Leaf
243	NPL-KIOM-133	Asteraceae	Artemisia	montana	Leaf
244	NPL-KIOM-134	Saururaceae	Houttuynia	cordata	Whole plant
245	NPL-KIOM-135	Oleaceae	Ligustrum	lucidum	Fruit and leaf
246	NPL-KIOM-136	Oleaceae	Ligustrum	japoricum	Fruit and leaf
247	NPL-KIOM-137	Oleaceae	Forsythia	viridissima	Fruit and leaf
248	NPL-KIOM-138	Oleaceae	Forsythia	suspensa	Fruit and leaf
249	NPL-KIOM-139	Nelumbonaceae	Nelumbo	nucifera	Seed and root
250	NPL-KIOM-140	Araliaceae	Acanthopanax	sessiliflorum	Root
251	NPL-KIOM-141	Scolopendridae	Scolopendra	subspinipes	Whole plant
252	NPL-KIOM-142	Schisandraceae	Schisandra	chinensis	Fruit
253	NPL-KIOM-143	Rutaceae	Evodia	rutaecarpa	Fruit and seed
254	NPL-KIOM-144	Lauraceae	Lindera	strichnifolia	Whole plant
255	NPL-KIOM-145	Convallariaceae	Polygonatum	odoratum	Root
256	NPL-KIOM-146	Gentianaceae	Gentiana	manshurica	Root
257	NPL-KIOM-147	Gentianaceae	Gentiana	scabra	Root
258	NPL-KIOM-148	Sapindaceae	Dimocarpus	longan	Fruit
259	NPL-KIOM-149	Asteraceae	Arctium	lappa	Root
260	NPL-KIOM-150	Amaranthaceae	Achyranthes	japonica	Root
261	NPL-KIOM-151	Zingiberaceae	Curcuma	wenyujin	Root
262	NPL-KIOM-152	Polygalaceae	Polygala	tenuifolia	Root
263	NPL-KIOM-153	Ranunculaceae	Clematis	manshurica	Root
264	NPL-KIOM-154	Ranunculaceae	Clematis	chinensis	Root
265	NPL-KIOM-155	Ulmaceae	Ulmus	macrocarpa	Rind
266	NPL-KIOM-156	Lauraceae	Cinnamomum	cassia	Rind
267	NPL-KIOM-157	Orobanchaceae	Cistanche	deserticola	Whole plant
268	NPL-KIOM-158	Ginkgoaceae	Ginkgo	biloba	Leaf
269	NPL-KIOM-159	Berberidaceae	Epimedium	koreanum	Whole plant
270	NPL-KIOM-160	Berberidaceae	Epimedium	sagittatum	Whole plant
271	NPL-KIOM-161	Berberidaceae	Epimedium	brevicornum	Whole plant
272	NPL-KIOM-162	Poaceae	Coix	lacryma-jovi	Seed
273	NPL-KIOM-163	Lamiaceae	Leonurus	japonicus	Whole plant
274	NPL-KIOM-164	Zingiberaceae	Alpinia	oxyphylla	Root
275	NPL-KIOM-165	Caprifoliaceae	Lonicera	japonica	Flower and leaf
276	NPL-KIOM-166	Araliaceae	Panax	ginseng	Root
277	NPL-KIOM-167	Asteraceae	Artemisia	capillaris	Leaf
278	NPL-KIOM-168	Apiaceae	Angelica	acutiloba	Root
279	NPL-KIOM-169	Apiaceae	Angelica	acutiloba	Root
280	NPL-KIOM-170	Boraginaceae	Lithospermum	erythrorhizon	Root
281	NPL-KIOM-171	Boraginaceae	Arnebia	euchroma	Root
282	NPL-KIOM-172	Boraginaceae	Arnebia	guttata	Root
283	NPL-KIOM-173	Lamiaceae	Perilla	frutescens	Leaf
284	NPL-KIOM-174	Lamiaceae	Perilla	frutescens	Leaf
285	NPL-KIOM-175	Lamiaceae	Perilla	frutescens	Leaf
286	NPL-KIOM-176	Lamiaceae	Perilla	frutescens	Leaf
287	NPL-KIOM-177	Asteraceae	Aster	tataricus	Whole plant
288	NPL-KIOM-178	Paeoniaceae	Paeonia	lactiflora	Root
289	NPL-KIOM-179	Polyporaceae	Polyporus	umbellatus	Mycelium
290	NPL-KIOM-180	Apiaceae	Angelica	decursiva	Root
291	NPL-KIOM-181	Apiaceae	Peucedanum	praeruptorum	Root
292	NPL-KIOM-182	Liliaceae	Fritillaria	thunbergii	Bulb
293	NPL-KIOM-183	Ranunculaceae	Aconitum	carmichaeli	Root
294	NPL-KIOM-184	Fabaceae	Gleditsia	sinensis	Seed and rind
295	NPL-KIOM-185	Polygonaceae	Rheum	undulatum	Root
296	NPL-KIOM-186	Poaceae	Phyllostachys	nigra	Stem
297	NPL-KIOM-187	Poaceae	Phyllostachys	bambusoides	Stem
298	NPL-KIOM-188	Rutaceae	Citrus	aurantium	Rind and fruit
299	NPL-KIOM-189	Rutaceae	Citrus	natsudaidai	Rind and fruit
300	NPL-KIOM-190	Solanaceae	Lycium	chinense	Fruit
301	NPL-KIOM-191	Solanaceae	Lycium	barbarum	Fruit
302	NPL-KIOM-192	Rhamnaceae	Hovenia	dulcis	Fruit
303	NPL-KIOM-193	Asparagaceae	Anemarrhena	asphodeloides	Root
304	NPL-KIOM-194	Amaranthaceae	Kochia	scoparia	Whole plant
305	NPL-KIOM-195	Rutaceae	Poncirus	trifoliata	Fruit and rind
306	NPL-KIOM-196	Rosaceae	Sanguisorba	officinalis	Root
307	NPL-KIOM-197	Rosaceae	Sanguisorba	officinalis	Root
308	NPL-KIOM-198	Orobanchaceae	Rehmannia	glutinosa	Root
309	NPL-KIOM-199	Gentianaceae	Gentiana	dahurica	Root
310	NPL-KIOM-200	Rutaceae	Citrus	reticulata	Rind and fruit
311	NPL-KIOM-201	Rutaceae	Citrus	unshiu	Rind and fruit
312	NPL-KIOM-202	Zygophyllaceae	Tribulus	terrestris	Fruit
313	NPL-KIOM-203	Plantaginaceae	Plantago	asiatica	Seed and whole plant
314	NPL-KIOM-204	Plantaginaceae	Plantago	depressa	Seed and whole plant
315	NPL-KIOM-205	Asteraceae	Atractylodes	chinensis	Root
316	NPL-KIOM-206	Asteraceae	Atractylodes	lancea	Root
317	NPL-KIOM-207	Apiaceae	Cnidium	officinale	Root
318	NPL-KIOM-208	Apiaceae	Ligusticum	chuanxiong	Root
319	NPL-KIOM-209	Araceae	Arisaema	amurense	Root
320	NPL-KIOM-210	Orchidaceae	Gastrodia	elata	Bulb
321	NPL-KIOM-211	Asparagaceae	Asparagus	cochinchinensis	Root
322	NPL-KIOM-212	Rutaceae	Citrus	reticulata	Rind and fruit
323	NPL-KIOM-213	Rutaceae	Citrus	unshiu	Rind and fruit
324	NPL-KIOM-214	Zingiberaceae	Amomum	tsao-ko	Seed
325	NPL-KIOM-215	Zingiberaceae	Alpinia	katsumadai	Seed and root
326	NPL-KIOM-216	Rubiaceae	Gardenia	jasminoides	Fruit
327	NPL-KIOM-217	Anacardiaceae	Rhus	verniciflua	Rind
328	NPL-KIOM-218	Lamiaceae	Lycopus	lucidus	Whole plant
329	NPL-KIOM-219	Alismataceae	Alisma	orientale	Root
330	NPL-KIOM-220	Asteraceae	Inula	helenium	Root
331	NPL-KIOM-221	Smilacaceae	Smilax	china	Root
332	NPL-KIOM-222	Convolvulaceae	Cuscuta	chinensis	Whole plant
333	NPL-KIOM-223	Rubiaceae	Morinda	officinalis	Root
334	NPL-KIOM-224	Asteraceae	Eupatorium	fortunei	Whole plant
335	NPL-KIOM-225	Asteraceae	Taraxacum	coreanum	Whole plant
336	NPL-KIOM-226	Asteraceae	Taraxacum	mongolicum	Whole plant
337	NPL-KIOM-227	Asteraceae	Taraxacum	officinale	Whole plant
338	NPL-KIOM-228	Piperaceae	Piper	longum	Fruit
339	NPL-KIOM-229	Polygonaceae	Polygonum	multiflorum	Root
340	NPL-KIOM-230	Asteraceae	Eclipta	prostrata	Whole plant
341	NPL-KIOM-231	Asteraceae	Artemisia	iwayomogi	Leaf
342	NPL-KIOM-232	Fabaceae	Albizzia	julibrissin	Flower and fruit
343	NPL-KIOM-233	Araliaceae	Kalopanax	pictus	Rind
344	NPL-KIOM-234	Apiaceae	Glehnia	littoralis	Root
345	NPL-KIOM-235	Rosaceae	Prunus	armeniaca	Fruit
346	NPL-KIOM-236	Rosaceae	Prunus	mandshurica	Fruit
347	NPL-KIOM-237	Cyperaceae	Cyperus	rotundus	Root
348	NPL-KIOM-238	Lamiaceae	Elsholtzia	ciliata	Whole plant
349	NPL-KIOM-239	Scrophulariaceae	Scrophularia	buergeriana	Whole plant
350	NPL-KIOM-240	Scrophulariaceae	Scrophularia	ningpoensis	Whole plant
351	NPL-KIOM-241	Geraniaceae	Geranium	thunbergii	Whole plant
352	NPL-KIOM-242	Papaveraceae	Corydalis	ternata	Root
353	NPL-KIOM-243	Papaveraceae	Corydalis	yanhusuo	Root
354	NPL-KIOM-244	Lamiaceae	Schizonepeta	tenuifolia	Whole plant
355	NPL-KIOM-245	Fabaceae	Trigonella	foenum-graecum	Seed
356	NPL-KIOM-246	Polygonaceae	Polygonum	cuspidatum	Root
357	NPL-KIOM-247	Asteraceae	Carthamus	tinctorius	Flower and seed
358	NPL-KIOM-248	Asteraceae	Carthamus	tinctorius	Flower and seed
359	NPL-KIOM-249	Lamiaceae	Scutellaria	baicalensis	Root
360	NPL-KIOM-250	Fabaceae	Astragalus	membranaceus	Root
361	NPL-KIOM-251	Fabaceae	Astragalus	membranaceus	Root
362	NPL-KIOM-252	Ranunculaceae	Coptis	chinensis	Root
363	NPL-KIOM-253	Ranunculaceae	Coptis	deltoidea	Root
364	NPL-KIOM-254	Ranunculaceae	Coptis	japonica	Root
365	NPL-KIOM-255	Ranunculaceae	Coptis	teeta	Root
366	NPL-KIOM-256	Rutaceae	Phellodendron	amurense	Rind
367	NPL-KIOM-257	Rutaceae	Phellodendron	chinense	Rind
368	NPL-KIOM-258	Asparagaceae	Polygonatum	sibiricum	Root
369	NPL-KIOM-259	Asparagaceae	Polygonatum	cyrtonema	Root
370	NPL-KIOM-260	Asparagaceae	Polygonatum	falcatum	Root
371	NPL-KIOM-261	Asparagaceae	Polygonatum	kingianum	Root
372	NPL-KIOM-262	Apiaceae	Foeniculum	vulgare	Seed
373	NPL-KIOM-263	Magnoliaceae	Magnolia	officinalis	Rind
374	NPL-KIOM-264	Magnoliaceae	Magnolia	obovata	Rind
375	NPL-KIOM-265	Pedaliaceae	Sesamum	indicum	Seed
376	NPL-NET-001	Ericaceae	Vaccinium	bracteatum	Leaves
377	NPL-NET-002	Apocynaceae	Cynanchum	wilfordii	Root
378	NPL-NET-003	Apocynaceae	Cynanchum	wilfordii	Root
379	NPL-NET-004	Apocynaceae	Cynanchum	wilfordii	Root
380	NPL-NET-005	Apocynaceae	Cynanchum	wilfordii	Root
381	NPL-NET-006	Apocynaceae	Cynanchum	wilfordii	Root
382	NPL-NET-007	Apocynaceae	Cynanchum	wilfordii	Root
383	NPL-NET-008	Apocynaceae	Cynanchum	wilfordii	Root
384	NPL-NET-009	Apocynaceae	Cynanchum	wilfordii	Root
385	NPL-NET-010	Apocynaceae	Cynanchum	wilfordii	Root
386	NPL-NET-011	Apocynaceae	Cynanchum	wilfordii	Root
387	NPL-NET-012	Apocynaceae	Cynanchum	wilfordii	Root
388	NPL-NET-013	Apocynaceae	Cynanchum	wilfordii	Root
389	NPL-NET-014	Apocynaceae	Cynanchum	wilfordii	Root
390	NPL-NET-015	Apocynaceae	Cynanchum	wilfordii	Root
391	NPL-NET-016	Apocynaceae	Cynanchum	wilfordii	Root
392	NPL-NET-032	Apiaceae	Ostericum	koreanum	Root
393	NPL-NET-033	Araliaceae	Aralia	cordata	Root
394	NPL-NET-034	Apiaceae	Angelica	dahurica	Root
395	NPL-NET-035	Apiaceae	Anthriscus	sylvestris	Root
396	NPL-NET-036	Apiaceae	Angelica	gigas	Root
397	NPL-NET-037	Apiaceae	Angelica	acutiloba	Root
398	NPL-NET-038	Apiaceae	Angelica	sinensis	Root
399	NPL-NET-039	Apiaceae	Levisticum	officinale	Root
400	NPL-NET-040	Apiaceae	Angelica	gigas	Root
401	NPL-NET-041	Apiaceae	Angelica	gigas	Root
402	NPL-NET-042	Asteraceae	Chrysanthemum	cinerariifolium	Flower
403	NPL-NET-043	Loganiaceae	Strychnos	-	Seed
404	NPL-NET-044	Caprifoliaceae	Valeriana	officinalis	Root
405	NPL-NPL-001	Ginkgoaceae	Ginkgo	biloba	Leaves
406	NPL-NPL-002	Fabaceae	Sophora	japonica	Flowers
407	NPL-NPL-003	Asteraceae	Senecio	jacobaea	-
408	NPL-NPL-004	Ranunculaceae	Aconitum	carmichaelii	Root
409	NPL-NPL-005	Ranunculaceae	Aconitum	carmichaelii	Root
410	NPL-NPL-006	Ephedraceae	Ephedra	sinica	Aerial parts
411	NPL-NPL-007	Rutaceae	Phellodendron	amurense	Bark
412	NPL-NPL-008	Rutaceae	Phellodendron	amurense	Bark
413	NPL-NPL-009	Ranunculaceae	Coptis	japonica	Root
414	NPL-NPL-010	Ulmaceae	Ulmus	davidiana	Bark
415	NPL-NPL-011	Araliaceae	Acanthopanax	sessiliflorum	Root and stem
416	NPL-NPL-012	Paeoniaceae	Paeonia	suffriticosa	Root
417	NPL-NPL-013	Paeoniaceae	Paeonia	suffriticosa	Root
418	NPL-NPL-014	Fabaceae	Glycyrrhiza	glabra	Root
419	NPL-NPL-015	Fabaceae	Glycyrrhiza	glabra	Root
420	NPL-NPL-016	Lamiaceae	Scutellaria	baicalensis	Root
421	NPL-NPL-017	Rubiaceae	Uncaria	rhynchophylla	Stem and hooks
422	NPL-NPL-018	Rubiaceae	Uncaria	tomentosa	Stem and hooks
423	NPL-NPL-019	Taxaceae	Taxus	chinensis	-
424	NPL-NPL-020	Fabaceae	Genista	monspesulana	Leaves
425	NPL-NPL-021	Fabaceae	Astragalus	membranaceus	Root
426	NPL-NPL-022	Solanaceae	Lycium	chinensis	Fruit
427	NPL-NPL-023	Amaranthaceae	Achyranthes	bidentata	Root
428	NPL-NPL-024	Amaranthaceae	Achyranthes	japonica	Root
429	NPL-NPL-025	Amaranthaceae	Cyathula	officinalis	Root
430	NPL-NPL-026	Aquifoliaceae	Ilex	aquifolium	Leaves
431	NPL-NPL-027	Malvaceae	Adansonia	digitata	Fruit and leaves
432	NPL-NPL-028	Campanulaceae	Codonopsis	pilosula	Root
433	NPL-NPL-029	Apocynaceae	Catharanthus	roseus	Leaves
434	NPL-NPL-031	Poaceae	Zea	mays	Seed
435	NPL-NPL-032	Asteraceae	Calendula	arvensis	Flower
436	NPL-NPL-033	Moraceae	Morus	alba	Leaves
437	NPL-NPL-034	Moraceae	Morus	alba	Leaves
438	NPL-NPL-037	Boraginaceae	Arnebia	euchroma	Root
439	NPL-NPL-038	Asteraceae	Synurus	deltoides	Leaves
440	NPL-NPL-039	Fabaceae	Millettia	saptholobi	Stem
441	NPL-NPL-040	Apocynaceae	Hoodia	gordonii	Stem
442	NPL-NPL-041	Lamiaceae	Orthosiphon	stamineus	Leaves
443	NPL-NPL-042	Polygalaceae	Polygala	tenuifolia	Root
444	NPL-NPL-043	Polygalaceae	Polygala	tenuifolia	Root
445	NPL-NPL-044	Apiaceae	Angelica	acutiloba	Root
446	NPL-NPL-045	Schisandraceae	Schisandra	chinensis	Fruit
447	NPL-NPL-046	Asteraceae	Chrysanthemum	indicum	Flower
448	NPL-NPL-047	Rhamnaceae	Zizyphus	jujuba	Fruit
449	NPL-NPL-048	Apiaceae	Ligusticum	striatum	Root
450	NPL-RDA-001	Araliaceae	Panax	ginseng	Root
451	NPL-RDA-002	Zingiberaceae	Zingiber	officinale	Root
452	NPL-RDA-003	Schisandraceae	Schisandra	chinensis	Fruit
453	NPL-RDA-004	Paeoniaceae	Paeonia	lactiflora	Root
454	NPL-RDA-005	Asteraceae	Atractylodes	macrocephala	Root
455	NPL-RDA-006	Asteraceae	Artemisia	iwayomogi	Leaves
456	NPL-RDA-007	Campanulaceae	Codonopsis	lanceolata	Root
457	NPL-RDA-008	Apiaceae	Angelica	gigas	Root
458	NPL-RDA-009	Apiaceae	Cnidium	officinale	Root
459	NPL-RDA-010	Asteraceae	Artemisia	argyi	Leaves
460	NPL-RDA-011	Araliaceae	Acanthopanax	senticosus	Root and stem
461	NPL-RDA-012	Amaranthaceae	Achyranthes	japonica	Root
462	NPL-RDA-013	Fabaceae	Glycyrrhiza	uralensis	Root
463	NPL-RDA-014	Cornaceae	Cornus	officinalis	Fruit
464	NPL-RDA-015	Zingiberaceae	Curcuma	longa	Rhizome
465	NPL-RDA-016	Fabaceae	Astragalus	membranaceus	Root
466	NPL-RDA-017	Asteraceae	Ixeris	strigosa	Leaves
467	NPL-RDA-018	Lamiaceae	Scutellaria	baicalensis	Root
468	NPL-RDA-019	Rosaceae	Rubus	coreanus	Fruit
469	NPL-RDA-020	Araliaceae	Aralia	continentalis	Root
470	NPL-RDA-021	Poaceae	Coix	lacryma-jobi	Seed
471	NPL-RDA-022	Campanulaceae	Platycodon	grandiflorum	Root
472	NPL-RDA-023	Orobanchaceae	Rehmannia	glutinosa	Root
473	NPL-RDA-024	Asteraceae	Carthamus	tinctorius	Flower
474	NPL-RDA-025	Asparagaceae	Liriope	platyphylla	Root
475	NPL-RDA-026	Polygonaceae	Polygonum	multiflorum	Root
476	NPL-RDA-027	Asparagaceae	Polygonatum	sibiricum	Root
477	NPL-RDA-028	Apocynaceae	Cynanchum	wilfordii	Root
478	NPL-SUB-001	Apiaceae	Notopterygium	-	Root
479	NPL-SUB-002	Apiaceae	Glehnia	littoralis	Root
480	NPL-SUB-003	Apiaceae	Glehnia	littoralis	Root
481	NPL-SUB-004	Zingiberaceae	Zingiber	-	Root
482	NPL-SUB-005	Zingiberaceae	Alpinia	officinarum	Root
483	NPL-SUB-006	Apiaceae	Ligusticum	chuanxiong	Root
484	NPL-SUB-007	Polypodiaceae	Drynaria	fortunei	Root
485	NPL-SUB-008	Cibotiaceae	Cibotium	barometz	Root
486	NPL-SUB-009	Polygonaceae	Fagopyrum	dibotryis	Root
487	NPL-SUB-010	Campanulaceae	Platycodon	grandiflorus	Root
488	NPL-SUB-011	Lamiaceae	Salvia	miltiorrhiza	Root
489	NPL-SUB-012	Lamiaceae	Salvia	miltiorrhiza	Root
490	NPL-SUB-013	Lamiaceae	Salvia	miltiorrhiza	Root
491	NPL-SUB-014	Lamiaceae	Salvia	miltiorrhiza	Root
492	NPL-SUB-015	Apiaceae	Angelica	sinensis	Root
493	NPL-SUB-016	Campanulaceae	Adenophora	remotiflora	Root
494	NPL-SUB-017	Euphorbiaceae	Euphorbia	pekinensis	Root
495	NPL-SUB-018	Polygonaceae	Rheum	palmatum	Root
496	NPL-SUB-019	Apiaceae	Cnidium	officinale	Root
497	NPL-SUB-020	Apiaceae	Angelica	pubescens	Root
498	NPL-SUB-021	Araliaceae	Aralia	cordata	Root
499	NPL-SUB-022	Eucommiaceae	Eucommia	ulmoides	Rind
500	NPL-SUB-023	Dioscoreaceae	Dioscorea	oposita	Root
501	NPL-SUB-024	Poaceae	Imperata	cylindrica	Root
502	NPL-SUB-025	Urticaceae	Boehmeria	nivea	Root
503	NPL-SUB-026	Asteraceae	Aucklandia	lappa	Root
504	NPL-SUB-027	Asteraceae	Saussurea	costus	Root
505	NPL-SUB-028	Menispermaceae	Stephania	tetrandra	Root
506	NPL-SUB-029	Asparagaceae	Ledebouria	seseloides	Root
507	NPL-SUB-030	Ranunculaceae	Pulsatilla	chinensis	Root
508	NPL-SUB-031	Vitaceae	Ampelopsis	radix	Root
509	NPL-SUB-032	Stemonaceae	Stemona	japonica	Root
510	NPL-SUB-033	Paeoniaceae	Paeonia	lactiflora	Root
511	NPL-SUB-034	Asclepiadaceae	Cynanchum	stauntonii	Root
512	NPL-SUB-035	Apiaceae	Angelica	dahurica	Root
513	NPL-SUB-036	Asteraceae	Atractylodes	macrocephala	Root
514	NPL-SUB-037	Asteraceae	Rhaponticum	uniflorum	Root
515	NPL-SUB-038	Zingiberaceae	Kaempferia	galanga	Root
516	NPL-SUB-039	Dipsacaceae	Dipsacus	asper	Root
517	NPL-SUB-040	Araliaceae	Panax	notoginseng	Root
518	NPL-SUB-041	Araliaceae	Panax	notoginseng	Root
519	NPL-SUB-042	Araliaceae	Panax	notoginseng	Root
520	NPL-SUB-043	Phytolaccaceae	Phytolacca	acinosa	Root
521	NPL-SUB-044	Polygonaceae	Polygonum	multiflorum	Root
522	NPL-SUB-045	Acoraceae	Acorus	tatarinowii	Root
523	NPL-SUB-046	Liliaceae	Ophiopogon	japonicus	Root
524	NPL-SUB-047	Dipsacaceae	Dipsacus	asper	Root
525	NPL-SUB-048	Apiaceae	Bupleurum	chinense	Root
526	NPL-SUB-049	Ranunculaceae	Anemone	raddeana	Root
527	NPL-SUB-050	Schisandraceae	Schisandra	chinensis	Fruit
528	NPL-SUB-051	Zingiberaceae	Curcuma	longa	Root
529	NPL-SUB-052	Asparagaceae	Polygonatum	sibiricum	Root
530	NPL-SUB-053	Amaranthaceae	Achyranthes	bidentata	Root
531	NPL-SUB-054	Amaranthaceae	Achyranthes	bidentata	Root
532	NPL-SUB-055	Amaranthaceae	Achyranthes	bidentata	Root
533	NPL-SUB-056	Polygalaceae	Polygala	tenuifolia	Root
534	NPL-SUB-057	Ranunculaceae	Clematis	chinensis	Root
535	NPL-SUB-058	Caryophyllaceae	Stellaria	media	Root
536	NPL-SUB-059	Araliaceae	Panax	ginseng	Root
537	NPL-SUB-060	Araliaceae	Panax	ginseng	Root
538	NPL-SUB-061	Araliaceae	Panax	ginseng	Root
539	NPL-SUB-062	Araliaceae	Panax	ginseng	Root
540	NPL-SUB-063	Asteraceae	Aster	tataricus	Root
541	NPL-SUB-064	Boraginaceae	Arnebia	euchroma	Root
542	NPL-SUB-065	Paeoniaceae	Paeonia	lactiflora	Root
543	NPL-SUB-066	Apiaceae	Anthriscus	sylvestris	Root
544	NPL-SUB-067	Asparagaceae	Anemarrhena	asphodeloides	Root
545	NPL-SUB-068	Asparagaceae	Anemarrhena	asphodeloides	Root
546	NPL-SUB-069	Asparagaceae	Anemarrhena	asphodeloides	Root
547	NPL-SUB-070	Asparagaceae	Anemarrhena	asphodeloides	Root
548	NPL-SUB-071	Rosaceae	Sanguisorba	officinalis	Root
549	NPL-SUB-072	Boraginaceae	Arnebia	euchroma	Root
550	NPL-SUB-073	Asteraceae	Atractylodes	macrocephala	Root
551	NPL-SUB-074	Apiaceae	Ligusticum	chuanxiong	Root
552	NPL-SUB-075	Apiaceae	Ligusticum	chuanxiong	Root
553	NPL-SUB-076	Apiaceae	Ligusticum	chuanxiong	Root
554	NPL-SUB-077	Ranunculaceae	Semiaquilegia	adoxoides	Root
555	NPL-SUB-078	Asparagaceae	Asparagus	cochinchinensis	Root
556	NPL-SUB-079	Orchidaceae	Gastrodia	elata	Root
557	NPL-SUB-080	Orchidaceae	Gastrodia	elata	Root
558	NPL-SUB-081	Smilacaceae	Smilax	china	Root
559	NPL-SUB-082	Smilacaceae	Smilax	china	Root
560	NPL-SUB-083	Caryophyllaceae	Pseudostellaria	heterophylla	Root
561	NPL-SUB-084	Caryophyllaceae	Pseudostellaria	heterophylla	Root
562	NPL-SUB-085	Smilacaceae	Smilax	china	Root
563	NPL-SUB-086	Cruciferae	Isatis	tinctoria	Root
564	NPL-SUB-087	Zingiberaceae	Zingiber	-	Root
565	NPL-SUB-088	Cyperaceae	Cyperus	rotundus	Root
566	NPL-SUB-089	Scrophulariaceae	Scrophularia	ningpoensis	Root
567	NPL-SUB-090	Poaceae	Phragmites	communis	Root
568	NPL-SUB-091	Polygonaceae	Polygonum	cuspidatum	Root
569	NPL-SUB-092	Scrophulariaceae	Picrorhiza	kurroa	Root
570	NPL-SUB-093	Fabaceae	Hedysarum	polybotrys	Root
571	NPL-SUB-094	Araliaceae	Panax	ginseng	Root
572	NPL-SUB-095	Araliaceae	Panax	ginseng	Root
573	NPL-SUB-096	Lamiaceae	Scutellaria	baicalensis	Root
574	NPL-SUB-097	Fabaceae	Astragalus	membranaceus	Root
575	NPL-SUB-098	Fabaceae	Astragalus	membranaceus	Root
576	NPL-SUB-099	Fabaceae	Glycyrrhiza	uralensis	Root
577	NPL-SUB-100	Fabaceae	Glycyrrhiza	uralensis	Root
578	NPL-SUB-101	Fabaceae	Glycyrrhiza	uralensis	Root
579	NPL-SUB-102	Fabaceae	Glycyrrhiza	uralensis	Root
580	NPL-SUB-103	Fabaceae	Glycyrrhiza	uralensis	Root
581	NPL-SUB-104	Fabaceae	Glycyrrhiza	uralensis	Root
582	NPL-SUB-105	Fabaceae	Glycyrrhiza	uralensis	Root
583	NPL-SUB-106	Fabaceae	Glycyrrhiza	uralensis	Root
584	NPL-SUB-107	Fabaceae	Glycyrrhiza	uralensis	Root
585	NPL-SUB-108	Fabaceae	Glycyrrhiza	uralensis	Root
586	NPL-SUB-109	Fabaceae	Glycyrrhiza	uralensis	Root
587	NPL-SUB-110	Fabaceae	Glycyrrhiza	uralensis	Root
588	NPL-SUB-111	Fabaceae	Glycyrrhiza	uralensis	Root
589	NPL-SUB-112	Fabaceae	Glycyrrhiza	uralensis	Root
590	NPL-SUB-113	Fabaceae	Glycyrrhiza	uralensis	Root
591	NPL-SUB-114	Fabaceae	Glycyrrhiza	uralensis	Root
592	NPL-SUB-115	Fabaceae	Glycyrrhiza	uralensis	Root
593	NPL-SUB-116	Fabaceae	Glycyrrhiza	uralensis	Root
594	NPL-SUB-117	Fabaceae	Glycyrrhiza	uralensis	Root
595	NPL-SUB-118	Fabaceae	Glycyrrhiza	uralensis	Root
596	NPL-SUB-119	Fabaceae	Glycyrrhiza	uralensis	Root
597	NPL-SUB-120	Fabaceae	Astragalus	membranaceus	Root
598	NPL-SUB-121	Fabaceae	Astragalus	membranaceus	Root
599	NPL-SUB-122	Fabaceae	Astragalus	membranaceus	Root
600	NPL-SUB-123	Fabaceae	Astragalus	membranaceus	Root
601	NPL-SUB-124	Fabaceae	Astragalus	membranaceus	Root
602	NPL-SUB-125	Fabaceae	Astragalus	membranaceus	Root
603	NPL-SUB-126	Fabaceae	Astragalus	membranaceus	Root
604	NPL-SUB-127	Fabaceae	Astragalus	membranaceus	Root
605	NPL-SUB-128	Fabaceae	Astragalus	membranaceus	Root
606	NPL-SUB-129	Fabaceae	Astragalus	membranaceus	Root
607	NPL-SUB-130	Fabaceae	Astragalus	membranaceus	Root
608	NPL-SUB-131	Fabaceae	Astragalus	membranaceus	Root
609	NPL-SUB-132	Fabaceae	Astragalus	membranaceus	Root
610	NPL-SUB-133	Fabaceae	Astragalus	membranaceus	Root
611	NPL-SUB-134	Fabaceae	Astragalus	membranaceus	Root
612	NPL-SUB-135	Fabaceae	Astragalus	membranaceus	Root
613	NPL-SUB-136	Fabaceae	Astragalus	membranaceus	Root
614	NPL-SUB-137	Fabaceae	Astragalus	membranaceus	Root
615	NPL-SUB-138	Fabaceae	Astragalus	membranaceus	Root
616	NPL-SUB-139	Fabaceae	Astragalus	membranaceus	Root
617	NPL-SUB-140	Fabaceae	Astragalus	membranaceus	Root
618	NPL-SUB-141	Fabaceae	Astragalus	membranaceus	Root
619	NPL-SUB-142	Fabaceae	Astragalus	membranaceus	Root
620	NPL-SUB-143	Fabaceae	Astragalus	membranaceus	Root
621	NPL-SUB-144	Zingiberaceae	Zingiber	officinale	Root
622	NPL-SUB-145	Acoraceae	Acorus	tatarinowii	Root
623	NPL-SUB-146	Polygonaceae	Rheum	palmatum	Root
624	NPL-SUB-147	Ranunculaceae	Cimicifuga	foetida	Root
625	NPL-SUB-148	Lauraceae	Cinnamomum	cassia	Branch
626	NPL-SUB-149	Fabaceae	Glycyrrhiza	uralensis	Root
627	NPL-SUB-150	Rosaceae	Prunus	persica	Seed
628	NPL-SUB-151	Zingiberaceae	Alpinia	officinarum	Root
629	NPL-SUB-152	Acoraceae	Acorus	calamus	Root
630	NPL-SUB-153	Zingiberaceae	Curcuma	longa	Root
631	NPL-SUB-154	Zingiberaceae	Curcuma	aromatica	Root
632	NPL-SUB-155	Zingiberaceae	Zingiber	officinale	Root
633	NPL-SUB-156	Zingiberaceae	Alpinia	officinarum	Root
634	NPL-SUB-157	Acoraceae	Acorus	calamus	Root
635	NPL-SUB-158	Zingiberaceae	Curcuma	longa	Root
636	NPL-SUB-159	Zingiberaceae	Curcuma	aromatica	Root
637	NPL-SUB-160	Zingiberaceae	Zingiber	officinale	Root
638	NPL-SUB-161	Zingiberaceae	Alpinia	officinarum	Root
639	NPL-SUB-162	Acoraceae	Acorus	calamus	Root
640	NPL-SUB-163	Zingiberaceae	Curcuma	longa	Root
641	NPL-SUB-164	Zingiberaceae	Curcuma	aromatica	Root
642	NPL-SUB-165	Zingiberaceae	Zingiber	officinale	Root
643	NPL-SUB-166	Zingiberaceae	Alpinia	officinarum	Root
644	NPL-SUB-167	Acoraceae	Acorus	calamus	Root
645	NPL-SUB-168	Zingiberaceae	Curcuma	longa	Root
646	NPL-SUB-169	Zingiberaceae	Curcuma	aromatica	Root
647	NPL-SUB-170	Zingiberaceae	Zingiber	officinale	Root
648	NPL-SUB-171	Zingiberaceae	Alpinia	officinarum	Root
649	NPL-SUB-172	Acoraceae	Acorus	calamus	Root
650	NPL-SUB-173	Zingiberaceae	Curcuma	longa	Root
651	NPL-SUB-174	Zingiberaceae	Curcuma	aromatica	Root
652	NPL-SUB-175	Zingiberaceae	Zingiber	officinale	Root
653	NPL-HBD-332	Apiaceae	Angelica	gigas	Root
654	NPL-HBD-334	Apiaceae	Angelica	gigas	Root
655	NPL-HBD-336	Apiaceae	Angelica	gigas	Stems and leaves
656	NPL-HBD-338	Apiaceae	Angelica	gigas	Stems and leaves

Results and discussion

Selection of extraction solvents for ¹H NMR measurements

The selection of extraction solvents was done with the aim of the study in mind, i.e., building a database of the metabolic profile of medicinal plants for various applications. This required the detection of as many metabolites as possible on one hand and the robustness of all the variables, ¹H resonances in this case. For this, all protons involved in hydrogen bonding, such as -OH need to be removed due to their great susceptibility to temperature, concentration or pH, that results in large variations This is achieved by using deuterated methanol, water or their mixtures. However, 100% deuterated water is not recommended due to signal broadening caused by large molecules (proteins and polysaccharides) that can be extracted and the low solubility of secondary metabolites in aqueous medium. Thus, the ¹H NMR analysis was conducted on samples treated with two solvents: CH₃OH-d₄ and CH₃OH-d₄-KH₂PO₄ buffer (1:1, v/v). with the expectation of detecting a broader range of metabolites.

This is exemplified in Fig. 1 that shows a very distinct metabolome of samples of the roots of Astragalus membranaceus according to the solvent. Using CH₃OH-d₄-KH₂PO₄ buffer (1:1, v/v), higher levels of isoflavonoids were detected, while flavonoids were more abundant in the CH₃OH-d₄ extracts (Fig. 1A). In the case of sugars, the level of sucrose and glucose, as well as saponins vary in the extracts (Fig. 1B and C). Thus, both extracts provided different chemical profiles and ¹H NMR data obtained from both extracts were combined to get a wider range of metabolites. This study establishes a standardized ¹H NMR database of medicinal plants and introduces a framework for large-scale metabolomic similarity profiling that enables chemotaxonomic context, quality surveillance, and identification of metabolically similar candidates.

Fig. 1.

¹H NMR spectra of Astragalus membranaceus radix in the range of δ 8.3–5.7 (A), δ 5.6–3.0 (B), δ 3.0–0.5 (C) measured in CH₃OH-d₄-KH₂PO₄ buffer in D₂O (pH 6.0) (spectra on the top) and CH₃OH-d₄ (spectra on the bottom). 1: H-2 of isoflavonoids, 2: B-ring of flavonoids and isoflavonoids, 3: A-ring of flavonoids and isoflavonoids, 4: H-2’ and H-6’ of kaempferol analogues, 5: H-3’ and H-5’ of kaempferol analogues, 6: H-1 of sucrose, 7: H-1 of α-glucose, 8: H-1 of β-glucose, 9: H-2’ of sucrose, 10: aspartic acid and asparagine, 11: methyls of saponins.

Classification or quality control of medicinal plants based on their chemotaxonomy (examples of Angelica, Glycyrrhiza, and Schisandra genera)

The NMR-based metabolomic database constructed in this study comprises 656 herbal samples, representing 113 plant families, 243 genera, and 299 species (Table 1). To evaluate the feasibility of chemotaxonomic classification using NMR profiling, we selected representative genera for detailed analysis based on medicinal relevance and intra-genus diversity. Among them, Angelica and Glycyrrhiza were chosen as model cases to examine whether metabolomic clustering reflects genus-level classification. NMR-based metabolomics has become a valuable tool in chemotaxonomy for chemotype-based classification and quality control. For instance, chemical profiling of Euphorbia species using NMR has been employed to investigate phylogenetic relationships, demonstrating its utility in identifying biomarkers associated with specific genetic or ecological traits²¹. However, despite its advantages, previous applications of NMR-based chemical profiling in chemotaxonomy have predominantly focused on limited single species or closely related genera. This narrow scope has limited its broader applicability across diverse plant taxa and left interspecies variability largely unexplored. Current NMR studies have primarily succeeded at the species level due to the presence of distinct metabolic markers. To determine whether NMR-based chemical profiling could be effectively applied at higher taxonomic levels, such as the genus, it is essential to account for the inherent variability among species. Expanding the scope to the genus level may require the integration of more advanced multivariate models and a broader dataset encompassing multiple species. This approach could help identify shared, stable markers while addressing environmental and genetic variability, which often obscure or hide genus-level distinctions. In this study, two representative genera, Angelica and Glycyrrhiza, were investigated to test the feasibility of this approach by determining whether they were classified closely in the clustering. For this application, the ¹H NMR data obtained after processing with bucketing (every 0.04 ppm) of data from both CH₃OH-d₄ and CH₃OH-d₄-KH₂PO₄ buffer (1:1, v/v) extracts, was analysed by a soft independent modelling cluster analogy (SIMCA) analysis, for which Angelica and Glycyrrhiza were classified separately and the distance of each sample to the cluster was then measured.

The global PCA model showed R2X(cum) = 0.859 and Q2(cum) = 0.474. The genus Angelica represents one of the most diverse groups of medicinal plants. Notably, the herbal medicine commonly known as Angelica refers to different species depending on the region: A. sinensis in China, A. acutiloba in Japan, and A. gigas in Korea. Figure 2 displays the clustering of Angelica samples based on numerical similarity. Similarity was evaluated using the SIMCA class model in PCA-class (R2X(cum) = 0.691 and Q2(cum) = 0.549), and the Distance to the Model in X-space (DModX) was calculated as a diagnostic measure of each sample’s distance from the class model. Samples with DModX values below the critical threshold of 1.27235 (p < 0.05) were considered to fit closely within the Angelica cluster. These samples are listed in (Table 2).

Fig. 2.

Distance to the model (DModX) values from the cluster of Angelica species to other samples. Angelica species are in red marked by *. The SIMCA class model yielded R2X(cum) = 0.691 and Q2(cum) = 0.549. Samples with DModX of each sample less than 1.0 are listed in (Table 2).

Table 2.

Distance to the model (DModX) values of Angelica species in the model of soft independent model of class analogy (SIMCA) for the Angelica class and the distance of non-Angelica samples having DModX < 1.27235 (Dcrit = 0.05).

No.	Sample code	Genus	Species	M2. DModX [2] (Norm), Weighted residuals
1	NPL-KHU-022	Angelica	gigas	0.572366
2	NPL-KHU-066	Angelica	acutiloba	0.615103
3	NPL-KIOM-041	Angelica	gigas	0.571566
4	NPL-KIOM-080	Angelica	dahurica	1.07631
5	NPL-KIOM-081	Angelica	dahurica	0.670917
6	NPL-KIOM-168	Angelica	acutiloba	0.525833
7	NPL-KIOM-169	Angelica	acutiloba	0.68044
8	NPL-KIOM-180	Angelica	decursiva	1.28192
9	NPL-NET-034	Angelica	dahurica	0.565141
10	NPL-NET-036	Angelica	gigas	0.540228
11	NPL-NET-037	Angelica	acutiloba	0.579614
12	NPL-NET-038	Angelica	sinensis	0.552658
13	NPL-NET-040	Angelica	gigas	0.656834
14	NPL-NET-041	Angelica	gigas	0.472327
15	NPL-NPL-044	Angelica	acutiloba	2.33752
16	NPL-RDA-008	Angelica	gigas	0.570281
17	NPL-SUB-015	Angelica	sinensis	0.491321
18	NPL-SUB-020	Angelica	pubescens	1.57637
19	NPL-SUB-035	Angelica	dahurica	1.29916
20	NPL-HBD-332	Angelica	gigas	0.935537
21	NPL-HBD-334	Angelica	gigas	0.961212
22	NPL-HBD-336	Angelica	gigas	1.67758
23	NPL-HBD-338	Angelica	gigas	1.11894

Outliers are marked and defined as samples with DModX values exceeding the 95% confidence threshold (Dcrit = 0.05) based on SIMCA analysis (DModX > 1.27235).

Interestingly, among the analysed samples, Ostericum koreanum - represented by NPL-NET-032 and NPL-KIOM-008 - exhibited DModXPS[2] values of 0.77189 and 0.55494, respectively. Both values fall below the critical threshold of 1.27235, indicating strong alignment with the Angelica cluster based on M2 component analysis. This numerical similarity suggests that, despite being classified separately, these samples share substantial metabolomic features with Angelica. This finding aligns with recent literature that has reclassified Ostericum koreanum as Angelica reflex²², highlighting the potential of this clustering approach to distinguish or associate plant species at the genus level based on numerical metabolomic similarity.

The results of the Glycyrrhiza species were different. The clustering result of the SIMCA class model (R2X(cum) = 0.933 and Q2(cum) = 0.763) showed that non-Glycyrrhiza uralensis samples generally exhibited higher DModX values, clearly distinguishing them from Glycyrrhiza uralensis in the DModX plot (Fig. 3). The corresponding values are provided in Table 3. These results support the utility of a genus-level SIMCA model as a practical screening tool for detecting samples that deviate from a reference class, thereby contributing to within-genus quality control, although this observation requires validation using a larger set of non-G. uralensis samples.

Fig. 3.

Distance to the model (DModX) values from the cluster of Glycyrrhiza species to other samples. Glycyrrhiza species are in red marked by *. The SIMCA class model yielded R2X(cum) = 0.933 and Q2(cum) = 0.763. Samples with DModX of each sample less than 1.0 are listed in (Table 3).

Table 3.

Distance to the model (DModX) values of Glycyrrhiza species in the model of soft independent model of class analogy (SIMCA) for the Angelica class and the distance of non-Glycyrrhiza samples having DModX < 1.28031 (Dcrit = 0.05).

No.	Primary ID	Genus	Species	M3. DModX [5] (Norm), Weighted residuals
1	NPL-KIOM-003	Glycyrrhiza	glabra	1.60065
2	NPL-KIOM-004	Glycyrrhiza	inflata	1.72332
3	NPL-KIOM-005	Glycyrrhiza	uralensis	2.47028
4	NPL-NPL-014	Glycyrrhiza	glabra	0.209195
5	NPL-NPL-015	Glycyrrhiza	glabra	0.583945
6	NPL-RDA-013	Glycyrrhiza	uralensis	0.582314
7	NPL-SUB-099	Glycyrrhiza	uralensis	1.12873
8	NPL-SUB-100	Glycyrrhiza	uralensis	0.77614
9	NPL-SUB-101	Glycyrrhiza	uralensis	0.781782
10	NPL-SUB-102	Glycyrrhiza	uralensis	0.967299
11	NPL-SUB-103	Glycyrrhiza	uralensis	1.02192
12	NPL-SUB-104	Glycyrrhiza	uralensis	0.870343
13	NPL-SUB-105	Glycyrrhiza	uralensis	0.695624
14	NPL-SUB-106	Glycyrrhiza	uralensis	0.980659
15	NPL-SUB-107	Glycyrrhiza	uralensis	1.10028
16	NPL-SUB-108	Glycyrrhiza	uralensis	1.05198
17	NPL-SUB-109	Glycyrrhiza	uralensis	0.764558
18	NPL-SUB-110	Glycyrrhiza	uralensis	0.786217
19	NPL-SUB-111	Glycyrrhiza	uralensis	0.553723
20	NPL-SUB-112	Glycyrrhiza	uralensis	0.747157
21	NPL-SUB-113	Glycyrrhiza	uralensis	0.582447
22	NPL-SUB-114	Glycyrrhiza	uralensis	0.954206
23	NPL-SUB-115	Glycyrrhiza	uralensis	0.611436
24	NPL-SUB-116	Glycyrrhiza	uralensis	0.710566
25	NPL-SUB-117	Glycyrrhiza	uralensis	0.680215
26	NPL-SUB-118	Glycyrrhiza	uralensis	0.779586
27	NPL-SUB-119	Glycyrrhiza	uralensis	0.622515

Outliers are marked and defined as samples with DModX values exceeding the 95% confidence threshold (Dcrit = 0.05), based on SIMCA analysis (DModX < 1.28031).

To complement the visual interpretation of the global PCA, we quantified local neighborhood agreement in the PCA score space (t[1]–t[2]) for the two case-study genera. For each sample, we identified its k nearest neighbors (k = 1 or 10) and calculated the percentage of samples whose nearest-neighbor set contained at least one sample sharing the same genus label. Within Angelica (n = 23), the 1-NN and 10-NN agreements were 34.8% and 69.6%, whereas within Glycyrrhiza (n = 27) they were 81.5% and 96.3%, respectively.

To further explore the chemotaxonomic capabilities of NMR-based metabolomics, Schisandra chinensis was analyzed to investigate how geographical origins impacted clustering and chemical profiles. This study incorporated five Schisandra samples collected from different locations: Korea (NPL-NPL-045, NPL-RDA-003, NPL-KIOM-042), China (NPL-SUB-050), and the Netherlands (a market sample). As illustrated in Fig. 4, the clustering results revealed that samples from Korea and China clustered closely together, exhibiting strong chemical similarity. The Schisandra sample obtained from the Netherlands formed a cluster far from the Korean and Chinese samples. This finding emphasizes the proposed potential of the clustering method to trace medicinal plants’ geographical origins. Even within the same species, differences in the origin can lead to distinct chemical profiles, a pattern that was also evident in the standardized ¹H NMR spectra, Fig. 5 provides a representative visualization of these origin-dependent spectral differences observed at the ¹H NMR level. Importantly, the samples from Korea and China exhibited only minor variability and remained grouped within the same primary cluster, further supporting the consistency within their shared origin. Conversely, the Netherlands sample deviated significantly, reflecting its different cultivation environment or processing methods.

Fig. 4.

Dendrogram of 45 principal components of 656 samples employed in the study. Branch colors indicate sample origin: Korea (red), China (green), and the Netherlands (blue). 1: Schisandra chinensis fruits collected from Korea (NPL-NPL-045, NPL-RDA-003, and NPL-KIOM-042) in red and China (NPL-SUB-050) in green, 2: Schisandra chinensis fruits collected from a market in the Netherlands in blue.

Fig. 5.

¹H NMR spectra (600 MHz, CH₃OH-d₄-KH₂PO₄ buffer (1:1, v/v, pH 6.0) of Schisandra chinensis fruits collected from Korea (A: NPL-KIOM-042, B: NPL-RDA-003, and C: NPL-NPL-045), from China (D: NPL-SUB-050), and from a market of the Netherlands (E).

Such chemical variability underscores the importance of evaluating therapeutic consistency and quality reliability, as differences in cultivation or harvesting environments can influence the chemical profile and potentially affect therapeutic activity and overall bioactivity. Taken together, these findings demonstrate that NMR-based metabolomics is effective not only for discriminating medicinal plants across taxonomic levels but also for detecting environmentally driven chemical variation within species. The consistent clustering of samples by geographical origin further supports the utility of this approach for origin tracking and quality assessment, providing a robust analytical basis for the scientific modernization of traditional medicines.

Comparison between domestic and imported medicinal plants and searching for domestic alternatives to replace imported products

While medicinal plants may share their name or traditional usage, their chemical compositions can vary significantly depending on their geographic origin, genetic variation, and environmental factors as discussed previously. This discrepancy poses challenges for the quality control and functional consistency in herbal medicine. Therefore, rather than relying solely on taxonomic identity or traditional nomenclature, metabolomic profiling might offer a robust solution for characterizing and comparing actual chemical profiles, enabling a more accurate substitution and standardization of medicinal resources. For example, while Taxus chinensis is primarily known for diterpenoid alkaloids such as taxane derivates, recent studies have demonstrated that its leaves contain a variety of other bioactive metabolites²³. This highlights the limitations of relying solely on known marker compounds and reinforces the need for a more holistic metabolomic approach. By profiling and clustering species based on their full metabolite profiles, this study provides a more robust framework for identifying functionally similar and potentially substitutable medicinal plants. In this context, ¹H NMR-based metabolomics is particularly advantageous, as it enables holistic and untargeted profiling of the entire metabolic landscape of a plant, providing a macroscopic view of species-specific chemical variation. Building on the robustness of NMR-based metabolomics for chemotaxonomy and quality control, this study further explores its potential for identifying alternative medicinal plants. The ability to compare and classify plants based on their chemical profiles provides a strong foundation for discovering substitutes for existing medicinal resources, especially in cases where sustainability, cost, or availability is a concern.

With this in mind, we compared Taxus and Viscum, two widely used but taxonomically distant medicinal plants, that are found to be metabolically close in ¹H NMR analysis. As illustrated in Fig. 6, the ¹H NMR spectra of Taxus chinensis leaves (NPL-NPL-019) and Viscum album stems (NPL-KIOM-020) revealed distinct yet overlapping chemical profiles. To assess whether this approach could be generalized across medicinal species, we extended the analysis to additional taxa. The analysis of Uncaria tomentosa stem barks (NPL-NPL-018), Uncaria rhynchophylla hooks (NPL-NPL-017), and Vaccinium bracteatum stems (NPL-NET-001) further demonstrated the utility of NMR profiling in the discovery of alternatives to existing medicinal plants, as shown in Fig. 7. The results highlight the value of NMR-based metabolomics in this context. By uncovering both shared and unique chemical features, this approach supports the rational selection of alternative resources based on functional and therapeutic equivalence, although further validation is needed to confirm this. Importantly, the metabolic patterns identified in this study provide a directional framework for future research on medicinal plants.

Fig. 6.

Comparison of ¹H NMR spectra (600 MHz, CH₃OH-d₄) Taxus chinensis leaves (NPL-NPL-019, A) and Viscum album stems (NPL-KIOM-020, B).

Fig. 7.

Comparison of ¹H NMR spectra (600 MHz, CH₃OH-d₄) Uncaria tomentosa stem barks (NPL-NPL-018, A), Uncaria rhynchophylla hooks (NPL-NPL-017, B), and Vaccinium bracteatum stems (NPL-NET-001, C).

To compare the value of NMR-based studies with other available analytical techniques, we conducted a molecular networking analysis using UHPLC-DAD-QToF-MS data via the GNPS platform to investigate structural similarities between Taxus chinensis and Viscum album. To improve annotation accuracy, ion-identity molecular networking (IIMN) was applied, integrating both precursor ion masses and MS/MS fragmentation data. This enabled the construction of a detailed molecular network (Fig. 8), that allowed the direct comparison and visualization of shared and unique chemical features of both species, consistent with previously reported strategies for comparative chemotaxonomic analysis that used GNPS-based MN platforms²⁴ .

Fig. 8.

Feature-based molecular networking (FBMN) of taxonomically distinct medicinal plants highlighting metabolite overlaps and functional similarity. The network generated from UHPLC-DAD-QToF-MS data of the Taxus chinensis and Viscum album using ion identity molecular networking (IIMN) shows shared molecular families, including a lactone-related cluster, observed across both species. Node labels correspond to feature IDs, and the corresponding precursor m/z, RT, library match information, and MSI confidence levels are provided in Supplementaray Table S2.

Several structurally annotated compounds with known or predicted bioactivities were detected among the shared nodes. Putative annotations included taccalonolide B^25,26, a microtubule-stabilizing agent structurally related to paclitaxel²⁷; atractylenolide III²⁸, known for its anti-inflammatory and gastrointestinal modulatory effects; and pterosin B²⁹, a phenolic compound with reported enzyme inhibition properties. Additionally, shared clusters contained nodes presumably corresponding to micheliolide³⁰, further suggesting overlapping metabolite families between the two species.

Notably, both Taxus chinensis and Viscum album have been independently reported to exhibit anticancer and immunomodulatory effects, albeit through different mechanisms. While paclitaxel from Taxus is a well-known chemotherapeutic agent targeting microtubule dynamics, Viscum album extracts - particularly mistletoe lectins and viscotoxins- are used in complementary cancer therapies for their immune-modulating and cytotoxic activities. The identification of network features putatively matching taccalonolide B³¹ and tricin³², both of which have documented anticancer properties, supports the hypothesis of functional similarity. This biochemical convergence not only reflects a structural overlap but also suggests a potential pharmacological parallel worthy of further investigation. While some of the shared compounds are well-characterized, others remain structurally significant yet functionally unverified. Therefore, although pharmacological validation was beyond the scope of this study, the observed metabolomic overlap offers a rational basis for future research into functional substitutability and therapeutic equivalence.

A second molecular networking analysis was performed on Uncaria tomentosa, Uncaria rhynchophylla, and Vaccinium bracteatum (Fig. 9). Despite notable differences in their taxonomic classification and traditional medicinal applications, these three medicinal plants exhibited a considerable overlap in their secondary metabolite profiles. Uncaria tomentosa, traditionally used in South American ethnomedicine for its immunomodulatory and anti-inflammatory effects, has been shown to exhibit corresponding pharmacological activities, largely attributed to oxindole alkaloids³³. Neutroprotective and anti-inflammatory properties have also been reported in modern studies for Uncaria rhynchophylla, long utilized in East Asian medicine for treating hypertension and neurological disorders³⁴. Vaccinium bracteatum, widely applied in East Asian practices for hepatoprotection and detoxification, has similarly been reported to possess antioxidant and liver-protective effects³⁵. These overlapping pharmacological profiles further support the rationale for exploring functional substitution among these botanically distinct species.

Fig. 9.

Comparative FBMN analysis of Uncaria tomentosa, Uncaria rhynchophylla, and Vaccinium bracteatum, highlighting shared alkaloid and polyphenol clusters. (A) Molecular networking revealed clusters in alkaloid and polyphenol features. (B) Enlarged view of a representative cluster showing features putatively annotated as uncarine C and structurally related polyphenolic glycosides detected across all three species. Highlighted nodes are traceable to their feature IDs and corresponding annotation details, including precursor m/z, RT, library match information, and MSI confidence levels, provided in Supplementaray Table S3.

Molecular networking revealed shared clusters containing features allegedly corresponding to key alkaloids, such as uncarine C³⁶, alongside polyphenolic glycosides structurally distributed across all three species. Among the shared features, uncarine C has demonstrated neuroprotective potential through antioxidant and calcium-channel modulating effects³⁷, while several glycosides structurally analogous to tricin are known to exert anticancer and anti-inflammatory activity^38–40. These observations suggest a degree of biosynthetic convergence, particularly in the production of indole alkaloids and phenolic derivatives. Additionally, Vaccinium bracteatum shared chemical features with Uncaria species, reinforcing the notion of a chemically interconnected profile despite botanical divergence. Taken together, these findings highlight the potential for functional substitution or therapeutic complementation among these species, based on shared metabolite structures and overlapping pharmacological potential. Further validation through bioactivity-guided assays may clarify the extent of their functional equivalence. All molecular-network nodes are linked to their feature IDs and corresponding precursor m/z, RT, library match information, and MSI confidence level, and the sample metadata used in GNPS are provided in Supplementary Table S2-S3.

These findings further support the utility of molecular networking in screening functionally relevant alternatives across traditionally distinct medicinal species. This comparative analysis illustrates how metabolomic fingerprints can support evidence-based selection of candidate substitutes, offering practical solutions for sustainability and compliance with international resource-sharing protocols. All considered, the molecular networks derived from UHPLC-DAD-QToF-MS data demonstrate that functionally relevant metabolite clusters are not strictly confined by taxonomy. This supports the macroscopic metabolic approach and suggests practical frameworks for identifying alternative sources of bioactive compounds with equivalent therapeutic potential.

Applications of ¹H NMR in multi-herbal mixture profiling

To gain a comprehensive understanding of plant metabolites, it is essential to adopt a broad analytical perspective. NMR spectroscopy offers such a perspective, enabling detailed insights into plant metabolomics. Notably, NMR can also be effectively applied to complex mixtures of medicinal herbs - an approach particularly relevant in traditional medicine, where multi-herbal formulations are commonly used as therapeutic solutions.

This study addresses a key question: can the emergent properties of these mixtures - often cited in traditional medicine - be analysed to elucidate the mechanisms underlying their synergistic effects? Furthermore, if one or more components exhibit significant chemical variability due to environmental or geographical factors, how might this influence the overall therapeutic efficacy of the mixture?

These questions form the basis of this investigation into the application of NMR-based metabolomics for understanding, characterizing, and potentially optimizing multi-herbal formulations.

Multi-herbal medicines have long been a cornerstone of traditional medical systems, often demonstrating synergistic effects that cannot be replicated by single components alone. In this study, we focus on Huanglian Jiedu Decoction (HJD), a classical prescription in East Asian medicine, known for its potent anti-inflammatory, antimicrobial, and detoxifying properties being prescribed traditionally for heat-related syndromes, infections, and inflammatory conditions. HJD comprises four primary herbs: Coptis chinensis, Phellodendron amurense, Scutellaria baicalensis, and Gardenia jasminoides. A broad spectrum of bioactive compounds, including alkaloids (e.g., berberine, palmatine) and flavonoids (e.g., baicalin, geniposide), which contribute to its therapeutic efficacy have been identified in HJD previously.

Despite its clinical relevance, the precise chemical interactions among the components of HJD remain poorly understood. Recent trends in drug development have emphasized the integration of traditional medicinal knowledge with modern analytical techniques. Multi-herbal formulations like HJD are increasingly viewed as templates for novel drug discovery, owing to their multitarget therapeutic potential. While most current studies focus on efficacy evaluation, there is a growing need to apply advanced methodologies to investigate the underlying mechanisms and discover new clinical applications.

In this context, the present study employs NMR-based metabolomics to unravel the chemical complexity of HJD. Using ¹H NMR fingerprints and PCA, we compared individual herbs and mixtures to evaluate mixture positioning in score space.

For this approach, ¹H NMR spectra were acquired for each individual herb and their combinations. Principal component analysis (PCA) was used to visualize the chemical relationships among samples. Figure 10 shows the PCA score plot of ¹H NMR data from Scutellaria baicalensis (S), Phellodendron amurense (P), Coptis japonica (C), Gardenia jasminoides (G), and their mixtures. Distinct clustering of individual plant extracts (black symbols) was observed, while the mixtures showed different clustering patterns. Blue symbols represent co-extracted mixtures, and red symbols represent mixtures prepared from individually extracted components. Interestingly, both preparation methods yielded similar results, suggesting that the chemical profile of the mixture remains consistent regardless of the extraction method.

Fig. 10.

Score plot of principal component analysis based on ¹H NMR data of Coptis japonica (C), Phellodendron amurense (P), Gardenia jasminoides (G), Scutellaria baicalensis (S), and their mixtures. Symbols in black are single plants, blue ones are co-extracted, and red ones are mixture of individually extracted ones.

This study demonstrated the efficacy of NMR-based metabolomics in distinguishing component herbs, assessing chemical interactions, and identifying potential synergistic effects in complex herbal mixtures. Furthermore, the ability to model and predict the behaviour of herbal formulations can be extended by comparing experimental mixtures with multivariate expectations derived from the constituent extracts, enabling a systematic assessment of mixture positioning in score space.

In addition to enabling the monitoring of changes in chemical profiles, NMR analysis yielded another noteworthy result through HJD analysis. Specifically, the NMR spectra of the individual medicinal plants revealed significant chemical similarities between Phellodendron amurense (P) and Coptis japonica (C). As shown in Fig. 11, both species contain benzylisoquinoline alkaloids, such as berberine, as a major component. Despite these similarities, the presence of additional, distinct compounds served to differentiate the two, highlighting the sensitivity of NMR in detecting subtle but meaningful variations in chemical composition. These results underscore that, even among chemically similar herbs, each plant contributes uniquely to traditional formulations through its characteristic metabolite profile. This distinction plays a critical role in optimizing the overall therapeutic effect of the prescription and highlights the importance of carefully considering the individual contributions of each component when interpreting or modifying traditional herbal formulas. Although their general chemical profiles may appear alike, variations such as the presence or concentration of specific bioactive compounds like berberine suggest that substitutions between herbs should be approached with caution to preserve the formulation’s intended efficacy. By capturing both the individual and emergent metabolic profiles of herbal mixtures, NMR-based analysis supports a deeper mechanistic understanding of synergy in traditional formulations and provides a scalable approach for evaluating multi-component therapeutics.

Fig. 11.

Comparison between ¹H NMR spectra of Coptis japonica (A) and Phellodendron amurense (B). 1: ¹H resonances of berberine. 2: ¹H resonances of chlorogenic acid.

The implications of this research extend beyond HJD. Applying this approach to other widely used traditional formulations may reveal common chemical markers or synergistic patterns that contribute to the development of generalized quality control methods. Ultimately, insights from this study contribute to the rational design of multi-component therapeutics.

Ultimately, to assess how the insights from this study may contribute to the rational design of multi-component therapeutics, we examined whether the multivariate profile of mixed extracts could be approximated from the ¹H NMR spectra of the individual component herbs (Fig. 12). A key question emerging from these results is whether the multivariate profile of mixed extracts can be reasonably approximated from the ¹H NMR spectra of the individual plant extracts. To evaluate this at the multivariate level, we performed PCA using the ¹H NMR spectra of the individual extracts and the corresponding mixed extracts. Figure 12 addresses this by comparing the PCA positioning of the mixed extract relative to the individual extracts based on their ¹H NMR profiles. The high degree of similarity observed in the PCA score space suggests that the mixture profile can be approximated from the constituent extracts under the tested conditions. This finding holds significant potential for streamlining the analysis of complex herbal formulations, reducing the need for exhaustive experimental testing while enabling more efficient prediction of mixture properties.

Fig. 12.

PCA score plot based on ¹H NMR data of Coptis japonica (C), Phellodendron amurense (P), Gardenia jasminoides (G), Scutellaria baicalensis (S), and their mixtures. Black symbols represent single-herb extracts; blue symbols, co-extracted mixtures; red symbols, mixtures prepared by combining individually extracted components; and green symbols, computed average profiles calculated from the constituent single-herb spectra.

However, the advantage of achieving a reliable chemical characterization of herbal mixtures extends further to practical applications within global regulatory frameworks, such as the Nagoya Protocol. This international agreement requires equitable benefit-sharing in the use of biological resources, increasing the pressure to identify and use domestically sourced medicinal plants. However, formulating effective substitutes is challenging when traditional remedies rely on complex combinations of herbs, where therapeutic efficacy arises not from a single component but from synergistic interactions.

In this study, NMR-based metabolomic profiling was applied to analyse multi-herb mixtures and compare their chemical signatures at the formulation level. This approach enabled the identification of candidate substitutes whose combined metabolic fingerprints align with those of established mixtures. For example, clustering analysis of Angelica and Glycyrrhiza species illustrated how multivariate methods can reveal interspecies chemical similarity allowing informed domestic substitution decisions. By evaluating both structural overlap and functional potential, mixture-level analysis supports more precise and evidence-based substitution without compromising therapeutic consistency.

This strategy not only addresses scientific and therapeutic considerations but also aligns with regulatory demands for sustainable sourcing. It provides a path forward for reducing dependence on imports, conserving biodiversity, and strengthening compliance with international protocols. Future research should refine this framework by characterizing inter-compound interactions and validating the clinical relevance of identified substitutes. Collectively, these findings support the development of sustainable, regulation-compliant herbal therapeutics and highlight the value of a macroscopic, database-driven approach to medicinal plant analysis.

However, the formulation-level comparisons presented here primarily support substitution-oriented prioritization rather than definitive equivalence testing. Because ¹H NMR fingerprints emphasize relatively abundant metabolites, low abundant bioactives and trace markers may be under-represented. In addition, neighborhood similarity and clustering outcomes are contingent on the current database composition and may not fully capture within-species variability associated with origin, season, processing, and chemotype. Finally, the “candidate substitutes” identified in this study reflect chemical similarity at the mixture level and require follow-up confirmation using orthogonal analytical approaches, including targeted assays and verification with authentic standards where needed, and ultimately pharmacological or clinical validation to substantiate therapeutic interchangeability.

Methods

Medicinal plant samples

A total of 656 herbal samples were obtained from Korea, China and the Netherlands, being generously provided by Kyung Hee University (Seoul, Korea), Korea Institute of Oriental Medicine (Daejeon, Korea), National Institute of Horticultural and Herbal Science (Eumseong, Chungcheongbuk-do, Korea), and SU BioMedicine (Leiden, The Netherlands). All the samples were identified by authors (Young Pyo Jang, Jeeyeoun Jung, Dae Young Lee, and Mei Wang). A detailed description of the samples is given in Table 1.

Sample preparation for ¹H NMR analysis using two solvent systems (CH₃OH-d₄-KH₂PO₄ buffer and CH₃OH-d₄)

For the ¹H NMR analysis, two solvent systems were prepared: (1) CH₃OH-d₄-KH₂PO₄ buffer (1:1, v/v) and (2) CH₃OH-d₄. The CH₃OH-d₄-KH₂PO₄ buffer (1:1, v/v), a 90 mM phosphate buffer solution containing 0.58 mM TMSP (internal standard), was prepared by dissolving 1.232 g of KH₂PO₄ in 100 ml of water and adjusting the pH of the buffer to 6.0 using 1.0 M NaOD. For CH₃OH-d₄, hexamethyldisiloxane (HMDSO) was added to CH₃OH-d₄ (0.418 mM) as an internal standard. For the ground-dried plant samples, 30 mg of ground sample were transferred into 2 mL microtubes using a spatula and 1 mL of either CH₃OH-d₄-KH₂PO₄ buffer (1:1, v/v) or CH₃OH-d₄ were added to each tube, and vortexed for 1 min at room temperature. The microtubes were then ultrasonicated for 20 min at room temperature to facilitate extraction and centrifuged at 17,000 g at room temperature for 20 min (to obtain a clear supernatant). An aliquot of 300 µL of the supernatant were then transferred to a 3 mm NMR tube. The HJD mixture was prepared by combining the four constituent plants in an equal mass ratio (7.5 mg each; total 30 mg). Additionally, all subset combinations of the four components were prepared using equal mass ratios while maintaining a total mass of 30 mg. All samples were extracted in 1 mL of solvent (30 mg/mL) under the same vortexing, ultrasonication, and centrifugation conditions before ¹H NMR analysis.

¹H NMR analysis

NMR measurements were performed on a Bruker Avance-III 600 MHz standard bore liquid-state NMR spectrometer with a 14.1 Tesla magnetic field. In this field, ¹H resonates at 600.13 MHz. A type of TCI H&F/C/N-D cryoprobe with Z gradient was used. 3 mm NMR tubes (Z112272) were purchased from Cortecnet and used for the experiments. The temperature was kept constant at 298 K. For internal locking, CH₃OH-d₄ was used. For each proton experiment, a 30-degree pulse of 2.64 msec at 5.5 W power with a fid resolution of 0.36 Hz, 64 scans with a relaxation delay of 1.5 s, and acquisition time of 2.7 s was used taking a total time of 5 min to complete the experiment. The water signal was suppressed using a pre-saturation method and low-power selective irradiation at 0.3 Hz H₂O at 4.87 ppm. Time domain data was transformed to the frequency domain by Fourier transformation with a window function of exponential function and a line broadening parameter set to 0.3 Hz for smoothing. The generated spectrums were manually phased, baseline corrected and calibrated to TMSP-d₄ at 0.0 ppm or HMDSO at 0.06 ppm.

¹H NMR data processing for multivariate data analysis

Bucketing using NMR data processing software facilitates the normalization of ¹H NMR spectra obtained from plant extracts prepared with two solvent systems -CH₃OH-d₄ and a mixture of CH₃OH-d₄ with KH₂PO₄ buffer (pH 6.0, 1:1, v/v). In this study, TopSpin software (version 3.7, Bruker BioSpin GmbH) was used for spectral processing. To enable comparison of the relative concentrations of extracted metabolites, total intensity normalization was applied to the spectral region from δ 10.0 to δ 0.3, using a bucket width of 0.04 ppm. Regions corresponding to residual solvent peaks - δ 3.28–3.24 for CH₃OH-d₄ and δ 5.0–4.7 for HDO - were excluded from analysis. The resulting bucketed data from each solvent (243 variables in total) were combined into a single data matrix for each sample (486 variables in total). The complete data matrix for all 656 samples is provided in the Supplementary Table S1.

Multivariate-and statistical analysis

Multivariate and statistical analyses were performed using SIMCA-P software (version 18.0.1, Sartorius) based on matrices derived from ¹H NMR data. The bucketed dataset was analysed by principal component analysis (PCA) and hierarchical clustering analysis (HCA) using PCA-reduced components. Model quality and robustness were evaluated using the cumulative fraction of explained variance (R2X(cum)) and the predictive ability (Q2(cum)) based on seven-fold internal cross-validation. To investigate genus- and species-related metabolic variation, soft independent modeling of class analogy (SIMCA) was applied by defining plant genus or geographical origin as PCA classes, depending on the specific analysis. In SIMCA, separate local PCA models are constructed for each predefined class, and the distance of each sample to its corresponding class model is calculated as the distance to the model (DModX). For species-level effects, DModX values were calculated by setting each plant genus as a PCA class. DModX values were log-transformed and used to screen potential outliers; however, no samples were excluded solely based on outlier diagnostics. All data were scaled using unit variance (UV) scaling.

UHPLC-DAD-QToF analysis

Selected samples were further analyzed using a UHPLC-DAD-QToF system (Ultimate 3000, Thermo Scientific) coupled with a QTOF-II mass spectrometer (Bruker) operating in positive ESI mode. Samples were extracted in 70% methanol, filtered, and diluted tenfold prior to injection. Chromatographic separation was achieved on a 2.1 × 150 mm Kinetex C₁₈ column (2.6 μm) with a 0.3 mL/min gradient flow. The gradient program was set using 0.1% formic acid in water (A) and 0.1% formic acid in acetonitrile (B). The gradient elution profile was as follows: 5% B to 95% B (16 min), 95% B (2 min), and 5% B (2 min) and the total run time was 20 min. The oven temperature was set at 40 °C. The sample injection volume was 1 µL. MS parameters included a capillary voltage of 4000 V, a drying gas temperature of 350 °C with a flow rate of 6 L/min, and a nebulizer gas pressure of 2.0 bar. Sodium formate was used for mass calibration. Spectra were recorded in positive ESI full-scan mode over an m/z range of 100–1650. No pooled QC samples or internal standards were included, as the LC–HRMS analysis was conducted for qualitative and comparative molecular networking purposes.

Molecular networking analysis

A molecular network was constructed using the molecular networking (MN) workflow available on the GNPS platform (https://gnps.ucsd.edu). Specifically, Feature-Based Molecular Networking (FBMN) was applied. Raw LC–HRMS data were converted to the .mzML format using the MSConvert tool within the ProteoWizard software suite and subsequently uploaded to GNPS for feature deconvolution and molecular networking analysis. Sample group information was incorporated during post-network analysis to support comparative interpretation of molecular clusters. The precursor ion mass tolerance was defined as 0.02 Da, and the fragment ion tolerance in MS/MS was also set at 0.02 Da. Molecular networks were established by filtering edges to retain only those with a cosine similarity score greater than 0.5 and at least four matched peaks. Additionally, connections between nodes were preserved in the network only if each node was within the top 10 most similar nodes of the other. The maximum allowable size for each molecular family was limited to 100 nodes, and edges with lower scores were progressively removed until each molecular family complied with this maximum size. Spectral matches within the network were compared against GNPS public spectral libraries. Library spectra underwent the same filtering procedures as the experimental input data. Matches between experimental spectra and library spectra were accepted only if they achieved a cosine score higher than 0.7 and had a minimum of four matched peaks. Finally, the molecular network was visualized using Cytoscape software (version 3.9.1), a widely adopted bioinformatics tool for network analysis.

Supplementary Information

Below is the link to the electronic supplementary material.

Author contributions

YHC conceptualized the study and designed the experiments. YHC and HKK developed the analytical methods. SS, ÖE, and HKK and conducted the experimental work. YHC and SS performed data analysis. HGJVM contributed to the refinement of statistical and multivariate analysis methods. JJ, YPJ, and DYL selected the Korean medicinal plants and verified their identification. MW selected and verified the identification of the Chinese medicinal plants. SS and YHC drafted the manuscript. HS and HSB reviewed and revised the analytical data. All authors contributed to writing and provided feedback on the manuscript. All authors have read and approved the final version.

Funding

This research was partially supported by the Unlocking Nature’s Pharmacy from Bogland Species (UNPBS) Project under grant number DOJProject209825, funded by the Department of Justice, Ireland. Dr. Jeeyoun Jung was supported by the Korea Institute of Oriental Medicine, Republic of Korea (grant number: KSN2312021). Dr. Mei Wang was partially supported by the Programma EFROWEST-Netherland 2021–2027 (grant number: 203876) and Hangzhou Ganzhicao Technology Co. Ltd (grant number: GanCao 025009).

Data availability

Data sets generated during the current study are available from the corresponding author on reasonable request.

Declarations

Competing interests

The authors declare no competing interests.