Welcome to Acta Agronomica Sinica,

Acta Agronomica Sinica ›› 2026, Vol. 52 ›› Issue (3): 735-745.doi: 10.3724/SP.J.1006.2026.51073

• CROP GENETICS & BREEDING·GERMPLASM RESOURCES·MOLECULAR GENETICS • Previous Articles     Next Articles

Establishment of a near-infrared reflectance spectroscopy model for predicting β-glucan content in naked barley grain

Li Ying(), Shi Xiao-Xu, Liu Hai-Cui, Shi Lyu, Xue Ya-Guang, Wei Ya-Feng()   

  1. Jiangsu Yanjiang Area Institute of Agricultural Sciences / Key Laboratory of Recycling Agriculture of Nantong City, Nantong 226001, Jiangsu, China
  • Received:2025-08-06 Accepted:2025-11-18 Online:2026-03-12 Published:2025-11-20
  • Contact: *魏亚凤, E-mail: w-yafeng@163.com E-mail:1185077523@qq.com;w-yafeng@163.com
  • Supported by:
    江苏省农业科技自主创新资金项(CX(24)3089);Jiangsu Agricultural Science and Technology Innovation Fund Project(CX(24)3089)

Abstract:

β-glucan, known for its roles in regulating blood sugar, lowering blood lipids, and exhibiting anti-tumor and antioxidant properties, is one of the most important quality indicators of naked barley. Establishing a near-infrared (NIR) prediction model for β-glucan content enables rapid and non-destructive evaluation of this trait in early-generation naked barley materials, which can significantly enhance breeding efficiency. In this study, NIR spectra were collected from 215 naked barley grain samples, and β-glucan content was determined using enzymatic methods. Based on spectral scanning and Monte Carlo cross-validation, 10 outlier samples were excluded. The remaining samples were divided into calibration and validation sets using the SPXY method. Fourteen preprocessing techniques, including normalization, first derivative, and second derivative, were applied to the spectral data. Partial least squares regression (PLSR), support vector machine (SVM), and principal component regression (PCR) were employed to construct prediction models for β-glucan content. Among these, the SVM model combined with the second derivative preprocessing showed the best predictive performance. To further optimize this model, six feature variable selection methods—successive projections algorithm (SPA), uninformative variable elimination (UVE), competitive adaptive reweighted sampling (CARS), variable iterative space shrinkage algorithm (VISSA), bootstrapping soft shrinkage (BOSS), and iteratively retained informative variables (IRIV) were evaluated. The CARS-SD-SVM model achieved the highest accuracy, with R2C, RMSEC, R2P, RMSEP, and RPD values of 0.859, 0.272, 0.877, 0.237, and 2.790, respectively. This model not only demonstrated excellent predictive performance but also high reliability, making it suitable for the rapid analysis and prediction of β-glucan content in naked barley grains. It holds great potential for application in naked barley breeding and quality assessment.

Key words: naked barley, β-glucan content, near infrared spectroscopy (NIRS), prediction model, feature variable selection

Table S1

β-glucan content of 215 naked barley resources used in this study"

序号No. 统一编号Uniform No. 名称
Name
产地
Region
β-葡聚糖含量
β-glucan content (%)
1 ZDM2187 戴帽元麦 Daimaoyuanmai 中国江苏 Jiangsu, China 3.79
2 ZDM2189 青元麦 Qingyuanmai 中国江苏 Jiangsu, China 3.08
3 ZDM2207 建湖三月黄 Jianhusanyuehuang 中国江苏 Jiangsu, China 4.25
4 ZDM2212 四棱元麦 Silengyuanmai 中国江苏 Jiangsu, China 3.14
5 ZDM2236 紫元麦 Ziyuanmai 中国江苏 Jiangsu, China 3.52
6 ZDM2249 六棱孔麦 Liulengkongmai 中国江苏 Jiangsu, China 3.89
7 ZDM2251 淮安狗尾 Huai’angouwei 中国江苏 Jiangsu, China 2.93
8 — 藏青3000 Zangqing 3000 中国西藏 Xizang, China 5.91
9 ZDM2259 洋辣子 Yanglazi 中国江苏 Jiangsu, China 3.40
10 ZDM2263 紫元麦 Ziyuanmai 中国江苏 Jiangsu, China 3.61
11 ZDM2280 南通铁麦 Nantongtiemai 中国江苏 Jiangsu, China 3.84
12 ZDM2289 江都四棱子 Jiangdusilengzi 中国江苏 Jiangsu, China 3.03
13 ZDM2290 太县洋六棱 Taixianyangliuleng 中国江苏 Jiangsu, China 3.18
14 ZDM2293 白六棱 Bailiuleng 中国江苏 Jiangsu, China 2.97
15 ZDM2295 句容六棱元麦 Jurongliulengyuanmai 中国江苏 Jiangsu, China 3.19
16 ZDM2302 武进元麦 Wujinyuanmai 中国江苏 Jiangsu, China 2.93
17 ZDM2311 键米大麦 Jianmidamai 中国江苏 Jiangsu, China 3.64
18 ZDM2313 常熟六棱元麦 Changshuliulengyuanmai 中国江苏 Jiangsu, China 4.06
19 ZDM2316 立新2号 Lixin 2 中国江苏 Jiangsu, China 4.05
20 ZDM2318 通麦3号 Tongmai 3 中国江苏 Jiangsu, China 4.32
21 ZDM8508 江苏元麦110号 Jiangsuyuanmai 110 中国江苏 Jiangsu, China 4.30
22 ZDM8515 江苏元麦117号 Jiangsuyuanmai 117 中国江苏 Jiangsu, China 3.35
23 ZDM8522 江苏元麦123号 Jiangsuyuanmai 123 中国江苏 Jiangsu, China 2.94
24 ZDM8523 江苏元麦124号 Jiangsuyuanmai 124 中国江苏 Jiangsu, China 3.92
25 ZDM8541 江苏元麦140号 Jiangsuyuanmai 140 中国江苏 Jiangsu, China 3.07
26 ZDM8542 江苏元麦141号 Jiangsuyuanmai 141 中国江苏 Jiangsu, China 3.23
27 ZDM8544 江苏元麦143号 Jiangsuyuanmai 143 中国江苏 Jiangsu, China 2.92
28 ZDM8546 江苏元麦145号 Jiangsuyuanmai 145 中国江苏 Jiangsu, China 4.23
29 ZDM8547 江苏元麦146号 Jiangsuyuanmai 146 中国江苏 Jiangsu, China 4.18
30 ZDM8553 江苏元麦151号 Jiangsuyuanmai 151 中国江苏 Jiangsu, China 3.94
31 ZDM8558 江苏元麦156号 Jiangsuyuanmai 156 中国江苏 Jiangsu, China 4.02
32 ZDM8559 江苏元麦157号 Jiangsuyuanmai 157 中国江苏 Jiangsu, China 3.14
33 ZDM8563 江苏元麦160号 Jiangsuyuanmai 160 中国江苏 Jiangsu, China 3.62
34 ZDM8565 江苏元麦162号 Jiangsuyuanmai 162 中国江苏 Jiangsu, China 3.33
35 ZDM8567 江苏元麦164号 Jiangsuyuanmai 164 中国江苏 Jiangsu, China 3.74
36 ZDM8588 江苏元麦31号 Jiangsuyuanmai 31 中国江苏 Jiangsu, China 3.21
37 ZDM8602 江苏元麦44号 Jiangsuyuanmai 44 中国江苏 Jiangsu, China 3.79
38 ZDM8606 江苏元麦48号 Jiangsuyuanmai 48 中国江苏 Jiangsu, China 2.99
39 ZDM8609 江苏元麦50号 Jiangsuyuanmai 50 中国江苏 Jiangsu, China 2.99
40 ZDM8610 江苏元麦51号 Jiangsuyuanmai 51 中国江苏 Jiangsu, China 3.60
41 ZDM8612 江苏元麦53号 Jiangsuyuanmai 53 中国江苏 Jiangsu, China 2.91
42 ZDM8613 江苏元麦54号 Jiangsuyuanmai 54 中国江苏 Jiangsu, China 3.13
43 ZDM8614 江苏元麦55号 Jiangsuyuanmai 55 中国江苏 Jiangsu, China 3.31
44 ZDM8622 江苏元麦62号 Jiangsuyuanmai 62 中国江苏 Jiangsu, China 3.05
45 ZDM8625 江苏元麦65号 Jiangsuyuanmai 65 中国江苏 Jiangsu, China 3.45
46 ZDM8639 江苏元麦78号 Jiangsuyuanmai 78 中国江苏 Jiangsu, China 3.21
47 ZDM8658 江苏元麦95号 Jiangsuyuanmai 95 中国江苏 Jiangsu, China 3.95
48 ZDM8663 江苏元麦9号 Jiangsuyuanmai 9 中国江苏 Jiangsu, China 3.01
49 ZDM09357 苏511 Su 511 中国江苏 Jiangsu, China 3.66
50 ZDM09466 苏71-75-12 Su 71-75-12 中国江苏 Jiangsu, China 3.92
51 ZDM09494 早芒8号 Zaomang 8 中国江苏 Jiangsu, China 3.00
52 — 昆仑12 Kunlun 12 中国青海 Qinghai, China 4.25
53 — 昆仑13 Kunlun 13 中国青海 Qinghai, China 4.50
54 — 昆仑14 Kunlun 14 中国青海 Qinghai, China 4.43
55 — 昆仑15 Kunlun 15 中国青海 Qinghai, China 4.69
56 — 柴青1号 Chaiqing 1 中国青海 Qinghai, China 5.06
57 — 甘垦5号 Ganken 5 中国甘肃 Gansu, China 4.67
58 — 苏裸麦2号 Suluomai 2 中国江苏 Jiangsu, China 3.42
59 — 通0914 Tong 0914 中国江苏 Jiangsu, China 3.67
60 — 紫元麦 Ziyuanmai 中国西藏 Xizang, China 5.06
61 ZDM01467 肚里黄 Dulihuang 中国青海 Qinghai, China 4.70
62 — 5171 中国西藏 Xizang, China 5.34
63 — 甘青4号 Ganqing 4 中国青海 Qinghai, China 3.91
64 — 东陈元麦 Dongchenyuanmai 中国江苏 Jiangsu, China 3.39
65 — 泰兴元麦 Taixingyuanmai 中国江苏 Jiangsu, China 3.52
66 ZDM00006 罗道庄大麦 Luodaozhuangdamai 中国北京 Beijing, China 3.00
67 ZDM03261 临海裸大麦 Linhailuodamai 中国浙江 Zhejiang, China 3.28
68 ZDM07528 尼泊尔青稞 Nepalqingke 尼泊尔 Nepal 3.25
69 ZDM07534 皮当蓝 Pidanglan 中国西藏 Xizang, China 3.57
70 ZDM07554 西嘎果那 Xigaguona 中国西藏 Xizang, China 4.38
71 ZDM07571 竹居青 Zhujuqing 中国西藏 Xizang, China 4.75
72 ZDM07573 色堆短芒 Seduiduanmang 中国西藏 Xizang, China 4.62
73 ZDM07610 沙尔兰 Sha’erlan 中国西藏 Xizang, China 4.27
74 ZDM07612 沃马紫 Womazi 中国西藏 Xizang, China 5.03
75 ZDM07613 沃马兰 Womalan 中国西藏 Xizang, China 4.85
76 ZDM07635 油松 Yousong 中国西藏 Xizang, China 4.00
77 ZDM07646 定结牧紫 Dingjiemuzi 中国西藏 Xizang, China 4.35
78 ZDM07648 孟德黄 Mengdehuang 中国西藏 Xizang, China 3.96
79 ZDM07738 雄马兰 Xiongmalan 中国西藏 Xizang, China 3.22
80 ZDM08086 二长四短芒白青稞 Erchangsiduanmangbaiqingke 中国青海 Qinghai, China 4.32
81 — 藏青2000 Zangqing 2000 中国西藏 Xizang, China 5.08
82 ZDM08177 红胶泥 Hongjiaoni 中国青海 Qinghai, China 3.89
83 ZDM08193 钩芒紫青稞 Goumangziqingke 中国青海 Qinghai, China 4.04
84 ZDM08217 黑青稞 Heiqingke 中国青海 Qinghai, China 3.89
85 ZDM08227 蓝青稞 Lanqingke 中国青海 Qinghai, China 4.25
86 ZDM08706 阿黑 Ahei 中国四川 Sichuan, China 4.15
87 ZDM08751 色黑瑞朗 Seheiruilang 中国四川 Sichuan, China 5.38
88 ZDM08784 日大多 Ridaduo 中国四川 Sichuan, China 4.20
89 ZDM08811 足木足大头六棱 Zumuzudatouliuleng 中国四川 Sichuan, China 4.32
90 ZDM08841 小青稞 Xiaoqingke 中国四川 Sichuan, China 4.35
91 ZDM08846 荞窝青稞 Qiaowoqingke 中国四川 Sichuan, China 4.73
92 ZDM09082 长钩紫大麦 Changgouzidamai 中国云南 Yunnan, China 4.31
93 ZDM09113 光芒黄大麦 Guangmanghuangdamai 中国云南 Yunnan, China 3.76
94 ZDM09653 阿212 A 212 中国四川 Sichuan, China 3.38
95 ZDM09772 藏830928 Zang 830928 中国西藏 Xizang, China 4.69
96 ZDM09802 藏840714 Zang 840714 中国西藏 Xizang, China 3.79
97 WDM00772 Jet 丹麦 Denmark 4.35
98 WDM02118 H.tetrast.L.V.viol 捷克斯洛伐克 Czechoslovakia 4.27
99 WDM03138 670 国外 Abroad 4.97
100 WDM03195 Tet 国外 Abroad 4.40
101 WDM03202 83 加拿大 Canada 4.55
102 WDM03227 Black Hulless 丹麦 Denmark 4.59
103 WDM03228 Black Hull-less C.I.596 美国 USA 3.99
104 WDM03230 Black Hull-less C.I.1027 美国 USA 4.58
105 WDM03237 Brachiki 墨西哥 Mexico 4.41
106 WDM03238 C1334 加拿大 Canada 4.27
107 WDM03263 Gaffel byg nϕgen 丹麦 Denmark 3.95
108 WDM03279 H.polyst.V.trif.C 丹麦 Denmark 4.06
109 WDM03280 H.riol.B.Sch 丹麦 Denmark 4.26
110 WDM03315 Lyallpur 丹麦 Denmark 4.76
111 WDM03361 SIBIRI“S” 墨西哥 Mexico 4.24
112 WDM03362 SNJCO“S” 墨西哥 Mexico 3.52
113 WDM03390 C 2734 加拿大 Canada 4.07
114 WDM03406 MANKER-OREGANO 墨西哥 Mexico 4.08
115 WDM03411 OREGANO 墨西哥 Mexico 3.46
116 WDM03419 Streng 叙利亚Syria 3.41
117 WDM05243 Japanische419 德国 Germany 4.36
118 WDM05418 Muanbori 日本 Japan 3.05
119 WDM05675 Scout 加拿大 Canada 4.09
120 WDM06222 四国裸46号 Siguoluo 46 日本 Japan 4.83
121 WDM06224 四国裸87号 Siguoluo 87 日本 Japan 4.00
122 WDM06225 四国裸88号 Siguoluo 88 日本 Japan 3.47
123 WDM06256 御岛裸 Yudaoluo 日本 Japan 3.48
124 WDM06281 Ⅲ-526 日本 Japan 4.39
125 WDM06288 Ⅲ-558 日本 Japan 4.02
126 WDM06289 Ⅲ-564 日本 Japan 3.69
127 WDM06290 Ⅲ-570 日本 Japan 4.89
128 — 藏青690 Zangqing 690 中国西藏 Xizang, China 4.69
129 — 冬青18 Dongqing 18 中国西藏 Xizang, China 4.79
130 — 2016-品比36 2016-pinbi 36 中国西藏 Xizang, China 3.56
131 — 喜马拉雅22 Ximalaya 22 中国西藏 Xizang, China 4.54
132 — 藏青25 Zangqing 25 中国西藏 Xizang, China 5.31
133 — 喜马拉雅19 Ximalaya 19 中国西藏 Xizang, China 5.04
134 — 12-10-556 中国西藏 Xizang, China 4.85
135 — 2016-品比42 2016-pinbi 42 中国西藏 Xizang, China 5.47
136 — INBON-5 黎巴嫩 Lebanon 4.00
137 — 藏青17 Zangqing 17 中国西藏 Xizang, China 4.67
138 ZDM08338 藏青320 Zangqing 320 中国西藏 Xizang, China 5.03
139 — INBON-12 黎巴嫩 Lebanon 3.28
140 — INBON-31 黎巴嫩 Lebanon 4.38
141 — INBON-45 黎巴嫩 Lebanon 4.33
142 — INBON-67 黎巴嫩 Lebanon 3.47
143 — 紫钩元麦 Zigouyuanmai 中国西藏 Xizang, China 6.07
144 — 藏青18 Zangqing 18 中国西藏 Xizang, China 6.98
145 — 盐食裸2号 Yanshiluo 2 中国江苏 Jiangsu, China 3.31
146 — 通麦6号 Tongmai 6 中国江苏 Jiangsu, China 3.22
147 — 苏裸麦1号 Suluomai 1 中国江苏 Jiangsu, China 3.48
148 — 通麦8号 Tongmai 8 中国江苏 Jiangsu, China 3.39
149 — 通0612 Tong 0612 中国江苏 Jiangsu, China 3.63
150 — 22DJ2 中国江苏 Jiangsu, China 4.26
151 — 通1409 Tong 1409 中国江苏 Jiangsu, China 3.04
152 ZDM09748 藏830187 Zang 830187 中国西藏 Xizang, China 5.06
153 ZDM09479 苏72-63 Su 72-63 中国江苏 Jiangsu, China 3.35
154 ZDM08790 泽盖紫青稞 Zegaiziqingke 中国四川 Sichuan, China 4.59
155 WDM02131 H.Viol B.8c H 丹麦 Denmark 3.90
156 ZDM09817 藏85品3872 Zang 85 pin 3872 中国西藏 Xizang, China 3.89
157 WDM03233 Black Hull-less C.I.2320 美国 USA 5.16
158 WDM03203 84 加拿大 Canada 4.56
159 WDM04080 CMB-74A-1796-B-6B-1Y-1B-1Y-1B-OY 墨西哥 Mexico 4.13
160 WDM06306 Ⅴ-193 日本 Japan 3.90
161 — 甘垦糯2号 Gankennuo 2 中国甘肃 Gansu, China 4.02
162 — 甘垦糯3号 Gankennuo 3 中国甘肃 Gansu, China 6.20
163 ZDM03920 长毛老麦 Changmaolaomai 中国贵州 Guizhou, China 4.05
164 ZDM03930 乌麦 Wumai 中国贵州 Guizhou, China 3.02
165 ZDM04247 嘎夏折通 Gaxiazhetong 中国西藏 Xizang, China 4.07
166 ZDM04606 白六棱 Bailiuleng 中国西藏 Xizang, China 4.51
167 ZDM04655 满都 Mandu 中国西藏 Xizang, China 4.10
168 ZDM04673 洛冬宗 Luodongzong 中国西藏 Xizang, China 4.84
169 ZDM05672 长芒白青稞 Changmangbaiqingke 中国西藏 Xizang, China 4.24
170 ZDM05886 白青稞 Baiqingke 中国西藏 Xizang, China 4.61
171 ZDM05889 白青稞 Baiqingke 中国西藏 Xizang, China 4.45
172 ZDM05990 白青稞 Baiqingke 中国西藏 Xizang, China 5.67
173 ZDM06029 冬青稞 Dongqingke 中国西藏 Xizang, China 3.85
174 ZDM06039 尼协 Nixie 中国西藏 Xizang, China 5.19
175 ZDM06213 达翁尼柱 Dawengnizhu 中国西藏 Xizang, China 3.44
176 ZDM06887 密穗钩芒 Misuigoumang 中国西藏 Xizang, China 5.66
177 ZDM07171 短芒青稞 Duanmangqingke 中国西藏 Xizang, China 4.38
178 ZDM07211 短芒黑青稞 Duanmangheiqingke 中国西藏 Xizang, China 4.53
179 ZDM07398 嘎夏 Gaxia 中国西藏 Xizang, China 4.37
180 ZDM07421 嘎胸 Gaxiong 中国西藏 Xizang, China 4.78
181 ZDM08383 吉塘青稞 Jitangqingke 中国西藏 Xizang, China 6.09
182 ZDM00477 秃头大麦 Tutoudamai 中国山东 Shandong, China 4.01
183 ZDM00500 米大麦 Midamai 中国山东 Shandong, China 3.91
184 ZDM00850 黄露仁 Huangluren 中国河南 Henan, China 3.30
185 ZDM00851 裸大麦 Luodamai 中国河南 Henan, China 3.03
186 ZDM00867 火灯芒 Huodengmang 中国河南 Henan, China 4.24
187 ZDM00875 三月黄 Sanyuehuang 中国河南 Henan, China 4.14
188 ZDM01313 桦南裸1号 Huananluo 1 中国黑龙江 Heilongjiang, China 3.03
189 ZDM03730 色查1号 Secha 1 中国四川 Sichuan, China 4.83
190 ZDM08405 紫青稞 Ziqingke 中国西藏 Xizang, China 5.16
191 ZDM08406 拉绥青稞 Lasuiqingke 中国西藏 Xizang, China 2.88
192 ZDM08691 裸大麦 Luodamai 中国湖北 Hubei, China 3.19
193 ZDM08870 通化马棚四月黄 Tonghuamapengsiyuehuang 中国四川 Sichuan, China 3.96
194 ZDM08893 合江大麦 Hejiangdamai 中国四川 Sichuan, China 3.13
195 ZDM08894 华阳大麦 Huayangdamai 中国四川 Sichuan, China 3.39
196 ZDM08933 南充米大麦 Nanchongmidamai 中国四川 Sichuan, China 3.50
197 ZDM09098 短钩裸大麦 Duangouluodamai 中国云南 Yunnan, China 3.84
198 ZDM09113 光芒黄大麦 Guangmanghuangdamai 中国云南 Yunnan, China 3.78
199 ZDM09135 宽颖大麦 Kuanyingdamai 中国云南 Yunnan, China 4.02
200 ZDM09167 温泉青稞 Wenquanqingke 中国云南 Yunnan, China 3.73
201 WDM00169 高拉输克罗凡 Gaolashukeluofan 罗马尼亚 Romania 4.83
202 WDM00171 加3027 Jia 3027 加拿大 Canada 4.29
203 WDM03141 BULK RESIST DESC-500B-OY 墨西哥 Mexico 4.10
204 WDM03504 70168 日本 Japan 4.24
205 WDM03748 Bermejo“S” 墨西哥 Mexico 4.69
206 WDM06287 Ⅲ-555 日本 Japan 4.30
207 — 黑青稞 Heiqingke 中国四川 Sichuan, China 3.46
208 — 白青稞 Baiqingke 中国四川 Sichuan, China 4.73
209 — 紫青稞 Ziqingke 中国四川 Sichuan, China 4.19
210 — 08-1325 中国西藏 Xizang, China 4.25
211 — 四棱白青稞 Silengbaiqingke 中国四川 Sichuan, China 3.99
212 — 六棱白青稞 Liulengbaiqingke 中国四川 Sichuan, China 4.12
213 — 紫壳青稞 Zikeqingke 中国四川 Sichuan, China 4.63
214 — 紫壳六棱紫青稞 Zikeliulengziqingke 中国四川 Sichuan, China 6.38
215 — 四棱白青稞 Silengbaiqingke 中国四川 Sichuan, China 6.18

Fig. 1

Near-infrared reflectance spectrometer DA7250 and small sample tray used in this experiment"

Fig. 2

Distribution of β-glucan chemical values in 215 naked barley germplasm resources"

Fig. 3

Original near-infrared spectra of 210 naked barley grain samples"

Fig. 4

Outlier detection results using the Monte Carlo cross-validation method"

Table 1

Statistics of β-glucan content in calibration and validation sets"

样本集
Sample set
样本数
Sample size
平均值
Mean (%)
最大值
Max. (%)
最小值
Min. (%)
极差
Range
标准偏差
SD
变异系数
CV (%)
校正集Calibration set 164 4.08 6.38 2.88 3.50 0.72 17.58
验证集Validation set 41 3.90 5.31 2.92 2.39 0.66 16.96
总样本Total set 205 4.04 6.38 2.88 3.50 0.71 17.52

Fig. 5

Preprocessed spectra of the calibration set using different methods a: raw spectra; b: normalization (NOR); c: first derivative (FD); d: second derivative (SD); e: multiplicative scatter correction (MSC); f: standard normal variate (SNV); g: Savitzky-Golay convolution smoothing (SG); h: Savitzky-Golay convolution smoothing first derivative (SGFD); i: Savitzky-Golay convolution smoothing second derivative (SGSD); j: detrend (DE); k: first derivative + multiplicative scatter correction (MSC); l: first derivative + standard normal variate (SNV); m: Savitzky-Golay convolution smoothing (SG) + multiplicative scatter correction (MSC); n: Savitzky-Golay convolution smoothing (SG) + standard normal variate (SNV); o: standard normal variate (SNV) + multiplicative scatter correction (MSC)."

Table 2

Comparison of prediction results using different modeling methods"

模型
Model
预处理
Pretreat
校正决定
系数R2C
校正标准
偏差RMSEC
外部验证决定系数R2P 验证标准
偏差RMSEP
相对分析
误差RPD
PLSR 原始光谱 RAW 0.657 0.419 0.798 0.309 2.143
归一化 NOR 0.678 0.406 0.799 0.298 2.224
一阶导数 FD 0.743 0.363 0.808 0.295 2.245
二阶导数 SD 0.781 0.335 0.779 0.311 2.128
多元离散校正 MSC 0.710 0.385 0.804 0.306 2.166
标准正态校正 SNV 0.723 0.376 0.807 0.300 2.204
SavitZky-Golay卷积平滑法 SG 0.652 0.422 0.804 0.307 2.157
SavitZky-Golay卷积平滑一阶求导 SGFD 0.811 0.311 0.727 0.362 1.830
SavitZky-Golay卷积平滑二阶求导 SGSD 0.816 0.307 0.461 0.520 1.274
去散射 DE 0.717 0.381 0.817 0.292 2.269
一阶导数+多元离散校正 FD+MSC 0.777 0.338 0.821 0.294 2.252
一阶导数+标准正态校正 FD+SNV 0.776 0.338 0.839 0.271 2.441
SavitZky-Golay卷积平滑+多元离散校正 SG+MSC 0.698 0.393 0.796 0.311 2.131
SavitZky-Golay卷积平滑+标准正态校正 SG+SNV 0.711 0.384 0.804 0.301 2.197
标准正态校正+多元离散校正 SNV+MSC 0.710 0.385 0.804 0.304 2.180
SVM 原始光谱 RAW 0.462 0.536 0.540 0.468 1.414
归一化 NOR 0.546 0.492 0.601 0.419 1.581
一阶导数 FD 0.780 0.338 0.826 0.281 2.359
二阶导数 SD 0.935 0.185 0.856 0.269 2.460
多元离散校正 MSC 0.928 0.194 0.217 2.932 0.226
标准正态校正 SNV 0.927 0.195 0.770 0.343 1.931
SavitZky-Golay卷积平滑法 SG 0.461 0.536 0.539 0.468 1.413
SavitZky-Golay卷积平滑一阶求导 SGFD 0.904 0.226 0.773 0.333 1.990
SavitZky-Golay卷积平滑二阶求导 SGSD 0.743 0.375 0.621 0.408 1.624
去散射 DE 0.614 0.452 0.715 0.351 1.887
一阶导数+多元离散校正 FD+MSC 0.786 0.334 0.814 0.557 1.188
一阶导数+标准正态校正 FD+SNV 0.788 0.333 0.812 0.287 2.305
SavitZky-Golay卷积平滑+多元离散校正 SG+MSC 0.902 0.225 0.204 2.783 0.238
SavitZky-Golay卷积平滑+标准正态校正 SG+SNV 0.902 0.225 0.763 0.346 1.915
标准正态校正+多元离散校正 SNV+MSC 0.928 0.194 0.768 0.346 1.913
PCR 原始光谱 RAW 0.096 0.680 0.102 0.641 1.032
归一化 NOR 0.319 0.590 0.429 0.506 1.307
一阶导数 FD 0.487 0.512 0.543 0.461 1.436
二阶导数 SD 0.714 0.382 0.641 0.401 1.650
多元离散校正 MSC 0.454 0.528 0.510 0.471 1.406
标准正态校正 SNV 0.453 0.529 0.510 0.470 1.409
SavitZky-Golay卷积平滑法 SG 0.096 0.680 0.102 0.641 1.032
SavitZky-Golay卷积平滑一阶求导 SGFD 0.708 0.387 0.683 0.388 1.705
SavitZky-Golay卷积平滑二阶求导 SGSD 0.880 0.248 0.507 0.521 1.271
去散射 DE 0.407 0.551 0.487 0.490 1.351
一阶导数+多元离散校正 FD+MSC 0.533 0.489 0.623 0.429 1.544
一阶导数+标准正态校正 FD+SNV 0.609 0.447 0.676 0.382 1.730
SavitZky-Golay卷积平滑+多元离散校正 SG+MSC 0.454 0.528 0.510 0.471 1.406
SavitZky-Golay卷积平滑+标准正态校正 SG+SNV 0.454 0.528 0.510 0.470 1.409
标准正态校正+多元离散校正 SNV+MSC 0.454 0.528 0.510 0.470 1.408

Fig. 6

Feature wavelength selection results based on six algorithms SPA: successive projections algorithm; UVE: uninformative variables elimination; CARS: competitive adaptive reweighted sampling; VISSA: variable iterative space shrinking algorithm; BOSS: bootstrapping soft shrinkage; IRIV: iteratively retained informative variables."

Table 3

Prediction results of SVM models based on different feature selection methods"

方法
Method
变量数
Variable number
校正决定系数R2C 校正标准偏差RMSEC 外部验证决定系数R2P 验证标准偏差RMSEP 相对分析误差RPD
FULL-SD-SVM 351 0.935 0.185 0.856 0.269 2.460
SPA-SD-SVM 22 0.786 0.334 0.772 0.321 2.060
UVE-SD-SVM 114 0.789 0.338 0.817 0.281 2.357
CARS-SD-SVM 28 0.859 0.272 0.877 0.237 2.790
VISSA-SD-SVM 115 0.818 0.311 0.822 0.287 2.304
BOSS-SD-SVM 23 0.809 0.316 0.817 0.294 2.249
IRIV-SD-SVM 47 0.883 0.250 0.803 0.316 2.097

Fig. 7

Process and variable distribution of CARS-based feature selection after second derivative preprocessing (a): variation in selected variable count; (b): changes in RMSECV; (c): regression coefficients of selected variables during the CARS process. RMSECV: root mean square error of cross-validation."

Fig. 8

Correlation between predicted and chemically measured β-glucan content in the validation set"

[1] 白羿雄, 郑雪晴, 姚有华, 等. 青稞种质资源表型性状的遗传多样性分析及综合评价. 中国农业科学, 2019, 52: 4201-4214.
Bai Y X, Zheng X Q, Yao Y H, et al. Genetic diversity analysis and comprehensive evaluation of phenotypic traits in hulless barley germplasm resources. Sci Agric Sin, 2019, 52: 4201-4214 (in Chinese with English abstract).
doi: 10.3864/j.issn.0578-1752.2019.23.002
[2] 侯殿志, 沈群. 我国29种青稞的营养及功能组分分析. 中国食品学报, 2020, 20: 289-298.
Hou D Z, Shen Q. Analysis of nutrition and functional components of 29 kinds of highland barley in China. J Chin Inst Food Sci Technol, 2020, 20: 289-298 (in Chinese with English abstract).
[3] Meints B, Hayes P M. Breeding naked barley for food, feed, and malt. Plant Breed Rev, 2019, 43: 95-119.
[4] Iaccarino N, Khakimov B, Mikkelsen M S, et al. Structurally different mixed linkage β-glucan supplements differentially increase secondary bile acid excretion in hypercholesterolaemic rat faeces. Food Funct, 2020, 11: 514-523.
doi: 10.1039/c8fo02507j pmid: 31833481
[5] Zhang J J, Zheng M J, Zhou L Y, et al. Oat β-glucan alleviates muscle atrophy via promoting myotube formation and suppressing protein degradation. J Sci Food Agric, 2023, 103: 6252-6262.
doi: 10.1002/jsfa.v103.13
[6] Wu Y S, Shao Y, Shao X Y, et al. Qingke β-glucan and Lactobacillus mitigate neuroinflammation and enhance cognitive function in an Alzheimer’s disease mouse model. Int J Biol Macromol, 2025, 319: 145427.
doi: 10.1016/j.ijbiomac.2025.145427
[7] Mcclear B V, Glennie-Holmes M. Enzymic quantification of (1 to 3) (1 to 4)-β-d-glucan in barley and malt. J Inst Brew, 1985, 91: 285-295.
doi: 10.1002/jib.1985.91.issue-5
[8] 胡美玲, 郅晨阳, 薛晓梦, 等. 单粒花生蔗糖含量近红外预测模型的建立. 作物学报, 2023, 49: 2498-2504.
doi: 10.3724/SP.J.1006.2023.24241
Hu M L, Zhi C Y, Xue X M, et al. Establishment of near-infrared reflectance spectroscopy model for predicting sucrose content of single seed in peanut. Acta Agron Sin, 2023, 49: 2498-2504 (in Chinese with English abstract).
doi: 10.3724/SP.J.1006.2023.24241
[9] 王若楠, 张颖星, 于筱菡, 等. 基于近红外快速检测技术的谷子淀粉多样性分析及模型构建. 作物学报, 2025, 51: 1757-1768.
doi: 10.3724/SP.J.1006.2025.44194
Wang R N, Zhang Y X, Yu X H, et al. Near-infrared spectroscopic evaluation of starch diversity and model construction in foxtail millet. Acta Agron Sin, 2025, 51: 1757-1768 (in Chinese with English abstract).
doi: 10.3724/SP.J.1006.2025.44194
[10] Xu Y H, Dong Y, Liu J M, et al. Combination of near infrared spectroscopy with characteristic interval selection for rapid detection of rice protein content. J Food Compos Anal, 2025, 137: 106995.
doi: 10.1016/j.jfca.2024.106995
[11] Schmidt J, Gergely S, Schönlechner R, et al. Comparison of different types of NIR instruments in ability to measure β-glucan content in naked barley. Cereal Chem, 2009, 86: 398-404.
doi: 10.1094/CCHEM-86-4-0398
[12] 杨焰婷, 江谧, 詹瑞玲, 等. 青稞特征营养成分含量快速检测模型的建立及优化. 食品工业科技, 2024, 45(18): 228-238.
Yang Y T, Jiang M, Zhan R L, et al. Development and optimization of a rapid detection model for characteristic nutrient content in highland barley. Sci Technol Food Ind, 2024, 45(18): 228-238 (in Chinese with English abstract).
[13] Li L L, Li L, Gou G Y, et al. Feasibility analysis of nondestructive detection of multiple parameters of highland barley by near-infrared spectroscopy. J Sci Food Agric, 2025, 105: 5194-5206.
doi: 10.1002/jsfa.v105.10
[14] 李鑫瑜, 侯名语, 崔顺立, 等. 红色和黑色花生籽仁黄酮含量近红外模型的构建. 中国农业科学, 2025, 58: 1284-1295.
doi: 10.3864/j.issn.0578-1752.2025.07.003
Li X Y, Hou M Y, Cui S L, et al. Construction of near infrared spectrometry model for flavonoids content of peanut with red and black testa. Sci Agric Sin, 2025, 58: 1284-1295 (in Chinese with English abstract).
doi: 10.3864/j.issn.0578-1752.2025.07.003
[15] Liu J M, Luo X, Zhang D J, et al. Rapid determination of rice protein content using near-infrared spectroscopy coupled with feature wavelength selection. Infrared Phys Technol, 2023, 135: 104969.
doi: 10.1016/j.infrared.2023.104969
[16] Galvão R K H, Araujo M C U, José G E, et al. A method for calibration and validation subset partitioning. Talanta, 2005, 67: 736-740.
doi: 10.1016/j.talanta.2005.03.025 pmid: 18970233
[17] Zhang W, Feng W W, Cai Z Q, et al. A deep one-dimensional convolutional neural network for microplastics classification using Raman spectroscopy. Vib Spectrosc, 2023, 124: 103487.
doi: 10.1016/j.vibspec.2022.103487
[18] Deng G M, Liu H G, Li J Q, et al. Rapid prediction of nucleosides content and origin traceability of Boletus bainiugan using Fourier transform near-infrared spectroscopy combined with chemometrics. Spectrochim Acta A Mol Biomol Spectrosc, 2025, 327: 125421.
doi: 10.1016/j.saa.2024.125421
[19] Khatun M S, Al Masum A, Islam M H, et al. Short wave-near infrared spectroscopy for predicting soluble solid content in intact mango with variable selection algorithms and chemometric model. J Food Compos Anal, 2024, 136: 106745.
doi: 10.1016/j.jfca.2024.106745
[20] Jiang H, Wang J N, Chen Q S. Comparison of wavelength selected methods for improving of prediction performance of PLS model to determine aflatoxin B1 (AFB1) in wheat samples during storage. Microchem J, 2021, 170: 106642.
doi: 10.1016/j.microc.2021.106642
[21] 李杨, 李翠玲, 王秀, 等. 高光谱成像的黄瓜病虫害识别和特征波长提取方法. 光谱学与光谱分析, 2024, 44: 301-309.
Li Y, Li C L, Wang X, et al. Identification of cucumber disease and insect pest based on hyperspectral imaging. Spectrosc Spectr Anal, 2024, 44: 301-309 (in Chinese with English abstract).
[22] Deng B C, Yun Y H, Cao D S, et al. A bootstrapping soft shrinkage approach for variable selection in chemical modeling. Anal Chim Acta, 2016, 908: 63-74.
doi: 10.1016/j.aca.2016.01.001
[23] 邵园园, 王永贤, 玄冠涛, 等. 基于高光谱成像的肥城桃品质可视化分析与成熟度检测. 农业机械学报, 2020, 51(8): 344-350.
Shao Y Y, Wang Y X, Xuan G T, et al. Visual detection of ssc and firmness and maturity prediction for Feicheng peach by using hyperspectral imaging. Trans CSAM, 2020, 51(8): 344-350 (in Chinese with English abstract).
[24] Li P, Jin Q T, Liu H M, et al. Determination of soluble solids content in loquat using near-infrared spectroscopy coupled with broad learning system and hybrid wavelength selection strategy. LWT, 2024, 206: 116570.
doi: 10.1016/j.lwt.2024.116570
[25] Zhang H L, Zhan B S, Pan F, et al. Determination of soluble solids content in oranges using visible and near infrared full transmittance hyperspectral imaging with comparative analysis of models. Postharvest Biol Technol, 2020, 163: 111148.
doi: 10.1016/j.postharvbio.2020.111148
[26] Lin C, Chen X, Jian L, et al. Determination of grain protein content by near-infrared spectrometry and multivariate calibration in barley. Food Chem, 2014, 162: 10-15.
doi: 10.1016/j.foodchem.2014.04.056 pmid: 24874350
[27] Han Z G, Cai S G, Zhang X L, et al. Development of predictive models for total phenolics and free p-coumaric acid contents in barley grain by near-infrared spectroscopy. Food Chem, 2017, 227: 342-348.
doi: S0308-8146(17)30074-2 pmid: 28274442
[28] Albanell E, Martínez M, De Marchi M, et al. Prediction of bioactive compounds in barley by near-infrared reflectance spectroscopy (NIRS). J Food Compos Anal, 2021, 97: 103763.
doi: 10.1016/j.jfca.2020.103763
[29] 李栓明, 郭银巧, 王克如, 等. 小麦籽粒蛋白质光谱特征变量筛选方法研究. 中国农业科学, 2015, 48: 2317-2326.
doi: 10.3864/j.issn.0578-1752.2015.12.004
Li S M, Guo Y Q, Wang K R, et al. Research on variable selection of wheat kernel protein content with near-infrared spectroscopy. Sci Agric Sin, 2015, 48: 2317-2326 (in Chinese with English abstract).
[30] Yu Y, Chai Y H, Li Z T, et al. Quantitative predictions of protein and total flavonoids content in Tartary and common buckwheat using near-infrared spectroscopy and chemometrics. Food Chem, 2025, 462: 141033.
doi: 10.1016/j.foodchem.2024.141033
[31] Xiong C H, She Y X, Jiao X, et al. Rapid nondestructive hardness detection of black highland barley kernels via hyperspectral imaging. J Food Compos Anal, 2024, 127: 105966.
doi: 10.1016/j.jfca.2023.105966
[32] 陈淼, 侯名语, 崔顺立, 等. 不同种皮颜色花生糖含量近红外模型的构建. 光谱学与光谱分析, 2022, 42: 2896-2902.
Chen M, Hou M Y, Cui S L, et al. Construction of near-infrared model of peanut sugar content in different seed coat colors. Spectrosc Spectr Anal, 2022, 42: 2896-2902 (in Chinese with English abstract).
[33] Yang Z Y, Cheng Z, Su P Y, et al. A model for the detection of β-glucan content in oat grain based on near infrared spectroscopy. J Food Compos Anal, 2024, 129: 106105.
doi: 10.1016/j.jfca.2024.106105
[1] HU Mei-Ling, ZHI Chen-Yang, XUE Xiao-Meng, WU Jie, WANG Jin, YAN Li-Ying, WANG Xin, CHEN Yu-Ning, KANG Yan-Ping, WANG Zhi-Hui, HUAI Dong-Xin, JIANG Hui-Fang, LEI Yong, LIAO Bo-Shou. Establishment of near-infrared reflectance spectroscopy model for predicting sucrose content of single seed in peanut [J]. Acta Agronomica Sinica, 2023, 49(9): 2498-2504.
[2] LI Ying, LIU Hai-Cui, SHI Lyu, SHI Xiao-Xu, HAN Xiao, LIU Jian, WEI Ya-Feng. Genetic diversity and population structure analysis of naked barley germplasm resources in Jiangsu province [J]. Acta Agronomica Sinica, 2023, 49(10): 2687-2697.
[3] LEI Yong, WANG Zhi-Hui, HUAI Dong-Xin, GAO Hua-Yuan, YAN Li-Ying, LI Jian-Guo, LI Wei-Tao, CHEN Yu-Ning, KANG Yan-Ping, LIU Hai-Long, WANG Xin, XUE Xiao-Meng, JIANG Hui-Fang, LIAO Bo-Shou. Development and application of a near infrared spectroscopy model for predicting high sucrose content of peanut seed [J]. Acta Agronomica Sinica, 2021, 47(2): 332-341.
[4] LU Xiao-Ping,LIU Dan-Dan,WANG Shu-Yan,MI Fu-Gui,HAN Ping-An,Lü Er-Suo. Genetic Effects and Heterosis Prediction Model of Sorghum  bicolor × S.sudanense Grass [J]. Acta Agron Sin, 2014, 40(03): 466-475.
[5] WANG Chun-Ping,GAO Li-Juan,PAN Zhi-Fen,NIMA Zha-Xi,TANG Ya-Wei,ZENG Xing-Quan,LI Qiao,DENG Guang-Bing,LONG Hai,YU Mao-Qun. Polymorphism of Starch Granule-Associated Proteins and 5′ Leader Sequence of GBSSI Gene in Indigenous Naked Barley (Hordeum vulgare L.) from Qinghai-Tibetan Plateau in China [J]. Acta Agron Sin, 2012, 38(07): 1148-1154.
[6] FENG Wei;YAO Xia;TIAN Yong-Chao;ZHU Yan;LIU Xiao-Jun;CAO Wei-Xing. Predicting Grain Protein Content with Canopy Hyperspectral Remote Sensing in Wheat [J]. Acta Agron Sin, 2007, 33(12): 1935-1942.
Viewed
Full text


Abstract

Cited

  Shared   
  Discussed   
No Suggested Reading articles found!