JP2019070548A - Method for preparing prediction formula for predicting brewing characteristics of brewing raw material grain, and method for producing grain varieties using prediction formula - Google Patents

Abstract

To provide new means for predicting the brewing characteristics of a raw material grain from data obtained from a small amount of raw material grain samples.SOLUTION: Provided is a method for preparing a prediction formula for predicting brewing characteristics when producing a brewed material by using a raw material grain having unknown brewing characteristics. In the method, one or more samples for metabolome analysis are prepared from each of the raw material grain samples, a plurality of sets of metabolome analysis data are acquired, and brewing characteristic data are acquired for each of the sample raw material grain samples. Then, the metabolome analysis data are subjected to multiple regression analysis for each objective variable using the explanatory variables and the brewing characteristic data as the objective variables. The prediction formula prepared by the method can be used, for example, for growing a new grain variety having desired brewing characteristics.SELECTED DRAWING: Figure 1

Description

  The present invention relates to a method of creating a prediction formula for predicting the brewing characteristics (brewing related analysis values) of desired brewing material grains, and a method of creating a new cereal cultivar using the prediction formula.

  As the properties of raw material grains are important in brewed products using various grains such as sake, miso and soy sauce, beer, whiskey, shochu etc., we will develop higher quality raw materials and evaluate the raw materials Various methods have been developed. However, evaluation methods of raw material grains are not sufficient and have various problems.

  For example, sake and shochu are fermented with water, rice and rice bran, and the properties of rice, which is the raw material cereals, are the workability in these production processes, the process of rice making and fermentation, and finally, sake.・ It greatly affects the quality of shochu. Therefore, in order to develop and cultivate higher quality raw material rice, various efforts are being made such as breeding of new raw material rice and improvement of the cultivation method. In recent years, movement to produce sake using local rice original raw material rice is prosperous, and it can be said that the need for breeding is increasing more and more. Among the raw rice varieties for sake, the most popular one is Yamada, which is why it is easy to use sake with excellent processability and high quality sake, and it is easy to use sake breweries and brewing techniques. Be However, although over 80 years have passed since the birth of Yamada, there is still a small amount of raw rice that exceeds this. The reason is that it is difficult to efficiently promote breeding and selection of raw rice varieties.

In the breeding of raw rice, the initial selection after mating has been carried out using the characteristics of cultivation characteristics, grain size of brown rice, and appearance of heart white as an index. By this selection index, it is mostly selected from 500-4,400 individuals to 10-399 individuals in the first individual selection, and it is mostly narrowed to several lines by the time the brew characteristics are examined (Non-Patent Document 1-8). Therefore, in recent years, efforts have been made to make selection more efficient by examining several indicators (rice milling characteristics, water absorbency, protein content, etc.) at an early stage of breeding (Non Patent Literatures 4, 7, 9, 10) . However, even in this case, it is the third selection two generations after the individual selection. The reason why selection by brewing characteristics is not performed from the early stage of breeding is the amount of brown rice necessary for analysis. For example, in order to investigate the characteristics of polished rice, brown rice of the order of 100 g (or even several tens of g) is required. To harvest this amount of brown rice, a sufficient amount of the target rice is selected. The cultivation area and the corresponding generation (years and months) to obtain the required amount of seed meal are required. Furthermore, in case of test brewing, even in small-scaled scale, brown rice of the order of kg is needed, and the cultivation area requires a larger space. In addition, the test rice milling machine and the corresponding rice making equipment, steamer, thermostatic bath etc. Facilities and sufficient space are also needed. For example, assuming that the raw rice yield is 500 kg / 10a, 300 g of brown rice is required for examining rice properties, 3 kg for small scale preparation and 3 kg for small scale preparation, and 300 kg for large preparation scale. . To investigate 1,000 varieties of raw rice, 600 m ² (tennis court (261 m ² ) for 2 sides) for investigation of rice grain characteristics, 6,000 m ^{2 for} small preparation scale test brewery (American football field (5,400 m ² ) A large surface area of 600,000 m ² (approximately 13 Tokyo Dome (46,755 m ² )) is required for large-scale trial brewing. In addition, in order to carry out the test brewing of a small preparation scale, a thermostatic bath of 45 m ² (27 帖, about 2 LDK) with n = 3 is necessary for 1,000 lines. In particular, it is necessary to finish the milling within a period in which the migration of minerals in brown rice is small, and even if 200 grams can be milled 70% in 15 minutes, it takes 3,000 hours for 1,000 lines. Even with 10 test mills, it will take about a month. In 50% milled rice, it takes about half a year because it takes six minutes, 90 minutes. In other words, it can be said that more efficient breeding of raw rice can be achieved if the amount of brown rice required for evaluation of brewing characteristics is the first bottleneck, and the brewing characteristics can be evaluated with fewer samples.

  In addition, the quality of raw rice is greatly influenced by the climate and soil environment of the cultivation area, the amount and composition of fertilization, and so on. Such cultivation environment is different every year, and therefore, the brewing characteristics of the raw material rice also fluctuate every year. Changes in the characteristics of the raw material rice greatly affect changes in physical characteristics and ingredients in the brewing process, and further, the ingredients and quality of the final product, the yield of the brew and the utilization of the raw material. However, in order to conduct a brewing test before manufacturing, monthly time is required, which is impossible in time. Therefore, before entering into a full-scale brewing season, it is necessary to quickly grasp the characteristics of the raw material rice and to adjust the brewing methods such as the rice milling method, the koji making method, and the fermentation method.

  Therefore, for the evaluation of the raw material rice, the raw material rice nationwide uniform analysis method (non-patent document 11) (hereinafter referred to as the sake-rice uniform analysis method) of the sake-rice research group is widely used. Since the analysis results by the sake-rice unified analysis method are related to the tendency of melting of raw rice, growth of Aspergillus oryzae, yeast, etc., they are used to estimate fermentation management and brew characteristics. However, there are 14 analysis items in the Sake-US unified analysis method, and it takes a lot of effort to evaluate the brewing characteristics. Therefore, there are attempts to predict other analysis values from analysis items that are relatively easy to analyze in this method (Non-Patent Document 12), and attempts to predict from completely different methods (Non-Patent Document 13), A prediction model limited to a specific analysis value and not accurate enough has not been built. Furthermore, in the analysis value by the liquor-rice unified analysis method, although the analysis value of the raw material rice itself is obtained, the characteristics of the brewed product during the brewing process, such as the course of the koji and the enzyme activity Can not predict. In other words, it can be said that there is much room for improvement in the evaluation method of raw material rice.

  In recent years, from data obtained by metabolomic analysis by GC-MS, LC-MS, UPLC-QTof-MS, etc., components to be included in the final product are evaluated by sensory evaluation score or exposition from projection to latent structures (PLS) regression analysis etc. A model has been reported that predicts the ranking of the. This is sake (non-patent document 14), soy sauce (non-patent document 15), wine (non-patent document 16), beer (non-patent document 17), Chinese rice wine (non-patent document 18), cheese (patent document 1). 2), cacao mass (patent document 3), coffee (patent document 4), and tea leaves (patent document 5). On the other hand, no model has been reported that predicts brewing characteristics, fermentation process, components and quality of final products by metabolome analysis of raw materials.

JP, 2013-7732, A JP 2015-55494 A JP, 2017-26558, A Unexamined-Japanese-Patent No. 2017-32326 JP, 2009-14700, A

Breeding of a new rice cultivar suitable for brewing sake, Mitsuru Kato, Naoki Sugiura, Takako Tsuji, Mitsuhiro Nakamura, Shirota Masahiro, Kato Takehiro, Takeo Funao, Satoru Kudo, Hiromi Kato, Akihiko Sawada, Toshio Suzuki, Toichiro Ichiro, Inoue Masakatsu, Yamamoto Koji, Ito Akitoshi, Aichi Agri-Study Report, 43, 17-24 (2011) Breeding progress of sake rice cultivar "Yamada rice cake" and mother cultivar "Yamada Sho", history of "Shokodate Tosen", Ikegami Masaru, Miyoshi Akihiro, Seko Harumi, Shibuya Ikuo, Nishida Kiyomi, Hyogo Agricultural Research Institute (Agriculture), 53, 37-50 (2005) Breeding progress of sake rice cultivar "Improved Yamada-Saki" and evaluation of brewing aptitude by the Suishu Research Group, Ikegami Masaru, Nishida Kiyokazu, Hyogo Agricultural Technology Research Institute (Agriculture), 55, 27-38 (2007) New rice cultivar suitable for rice cultivar "Etsubari", Kazuhiko Ishizaki, Kazuyuki Kobayashi, Takaya Matsui, Satoshi Kanada, Toyoda Hoshi, Yukio Sasaki, Toshifumi Abe, Tokubun Abe, Keiichi Abe, Keiko Higuchi, Hironobu Shigeyama, Hirao Kenichi, Niigata Prefectural Agricultural Research Institute Research Report, 9, 81-87 (2008) A New Rice Variety “Fukunohana”, Tetsuya Hirano, Dr. Uchiyamada, Koji Kondo, Hirotaka Maeda, Akira Matsumoto, Yoshihiro Akama, Bulletin of Tohoku Agricultural Research Station Research Bulletin, 7, 9-16 (1967) On the breeding of a new rice cultivar suitable for sake making, “Hatore No. 1”, Masashi Maejima, Toka Hisaka, Edo Yoshiharu, Takihiro Tokuo, Hiroshima Prefectural Agricultural Experiment Station Report, 48, 1-6 (1984) Breeding of new rice cultivar "Yamagata Sake No. 49", Hiroshi Shibata, Sato Koichi, Kikuchi Eiichi, Nakaba Masato, Sano Tomoyoshi, Tanifuji Yuji, Nakaba Rieko, Kuroki Ikuo, Yokoo Nobuhiko, Yukio Kazuhiro, Goto Kiyozo , Yamakawa Minoru, Yamagata Prefectural Agricultural Experiment Station Research Report, 30, 1-15 (1996) Breeding of new rice varieties "Autumn spirit", Shibasaki Takumi, Kato Takemitsu, Kashiyama Toshihiko, Matsumoto Shinichi, Kawamoto Yoshihiko, Yamamoto Tatsuhiko, Shizukuishi Susumu, Saito Shoichi, Fukuda Kenshiro, Shikinuki Kazuo, Ikeda Naomi, Akita agriculture Laboratory research report, 40, 23-43 (1999) Development of rice cultivar rice variety "Kotan Rei" and sake rice breeding in Niigata prefecture, Kobayashi Kazuyuki, Kanada Satoshi, Matsuda Takaya, Ishizaki Kazuhiko, Nabekura Yoshihito, Watanabe Kenichi, Breeding Research, 8, 55 -61 (2006) Development and sake brewing characteristics of sake brewing suitable rice new breed "Akita sake komachi", Takahashi Hitoshi, Taguchi Takanobu, Hokei, 98, 598-609 (2003) Sake-rice Study Group ed .: "Rate-wide integrated analysis method for raw materials for sake brewing", Sake-rice study group (1996) Studies on the suitability of rice for sake. (11)-On digestibility of steamed rice and prediction of digestion residue of steamed rice and digestion residue of rice bran, Hideo Hanamoto, Hidego Nakazawa, Seizo Takemura, Jyokyo, 73, 580-581 (1978) The relationship between meteorological data and sake brewing suitability of raw material rice, Masao Okuda, Katsumi Hashizume, Mikuyo Numata, Midori Upuse, Nami Goto, Shigeaki Mikami, Keikyo, 104, 699-711 (2009) Gas chromatography / mass spectrometry based component profiling and quality prediction for Japanese sake, Natsuki Mimura, Atsuko Isogai, Kazuhiro Iwashita, Takeshi Bamba, and Eiichiro Fukushima, Journal of Bioscience and Bioengineering, 118, 4, 406-414 (2014) Metabolite profiling of soy sauce using gas chromatography with time-of-flight mass spectrometry and analysis of correlation with quantitative descriptive analysis, Shinya Yamamoto, Takeshi Bamba, Atsushi Sano, Yukako Kodama, Miho Imamura, and Bioengineering, 114, 2, 170-175 (2012) Wine Metabolomics: Objective Measures of Sensory Properties of Semillon from GC-MS Profiles, Leigh M. Schmidtke, John W. Blackman, Andrew C. Clark, and Paris Grant-Preece, J. Agric. Food Chem., 61, 49, 11957 -11967 (2013) Predictive analysis of beer quality by correlating sensory evaluation with higher alcohol and ester production using multivariate statistics methods, Jian-Jun Dong, Qing-Liang Li, Hua Yin, Cheng Zhong, Jun-Guang Hao, Pan-Fei Yang, , Shi-Ru Jia, Food Chemistry, 161, 15, 376-382 (2014) Intelligent sensing sensory quality of Chinese rice wine using near infrared spectroscopy and nonlinear tools, Qin Ouyang, Quansheng Chen, Jiewen Zhao, Spectrochimica Acta Part A: Molecular and Biomolecular Spectroscopy, 154, 5, 42-46 (2016)

  An object of the present invention is to provide a novel means that makes it possible to predict the brewing characteristics of raw material cereals from data obtained from small amounts of raw material cereal samples.

  The inventors of the present invention developed a method of constructing a prediction formula for predicting the brewing characteristics of the raw material rice from metabolomic analysis data of the raw material rice as a result of keen research, The present invention has been completed by finding that it is applicable to raw material grains of brewed products.

That is, the present invention is a method of creating a brewing characteristic prediction equation for predicting brewing characteristics when producing brews using raw material cereals for raw material cereals of unknown brewing characteristics,
Preparing one or more samples for metabolomic analysis from each of a plurality of brew raw material cereal samples;
Using the metabolome analysis sample, metabolomic analysis is conducted to comprehensively analyze compounds contained in raw material grains, and metabolome analysis data composed of quantitative numerical data on the abundance of each compound in raw material grains is used. Acquiring process;
Obtaining at least one brewing characteristic data related to brewing characteristics for each of said brew material cereal samples;
Performing multiple regression analysis for each objective variable using two or more quantitative numerical data selected from each metabolome analysis data as an explanatory variable, and using a brewery characteristic data as an objective variable;
To provide a method, including:
Further, the present invention provides a step of preparing a brewing characteristic prediction equation by the method of the present invention described above;
Obtaining a cereal sample from a growing population of cereals or a population of cereal lines, preparing a sample for metabolomic analysis from the sample, and acquiring metabolomic analysis data;
Substituting the acquired metabolomic analysis data into a brewing characteristic prediction equation to calculate a prediction value of the brewing characteristic;
Selecting from the population individuals or lines predicted to have desired brewing characteristics based on the predicted value;
Provide a method of producing cereal varieties, including

  According to the present invention, metabolome analysis data of raw material rice, characteristic analysis value of raw material rice itself, enzyme activity of rice bran, characteristics during brewing such as fermentation process, components of moromi which is final product or distillation raw material, final brew For the first time, a prediction model that can predict components and quality is provided. By using the prediction formula prepared by the method of the present invention, with regard to a raw material rice line whose brewery related analysis value is unknown, the brewery related analysis value (evaluated by the liquor-rice unified analysis method from metabolomic analysis data of the raw material rice of that line) Characteristics of raw material rice, enzyme titer of persimmon, moromi course, characteristics of sake made). In metabolomic analysis, sufficient data can be obtained even from a small amount of raw material rice sample of about 0.1 g, so that it becomes possible to predict and evaluate the brew characteristics at an early stage of the raw rice breeding process, and the raw rice breeding Contribute significantly to speeding up Moreover, the prediction equation obtained by the present invention can be applied not only to an index of selection for breeding but also to an index of examination of a brewing method and fermentation management. Furthermore, the present invention is not limited to rice and sake, but also for raw material grains of various brews such as shochu, beer, whiskey, soy sauce, and miso, such as barley malt, wheat, soybean, etc. It can be created by the method.

It is a figure explaining the outline of the prediction equation creation method by the present invention. It is a figure explaining the actual work procedure for prediction expression creation.

  The present invention is a method of creating a brewing characteristics prediction equation for predicting brewing characteristics when producing brews using raw material cereals for raw material cereals of unknown brewing characteristics. The term "raw grain with unknown brewery characteristics" is a well-known variety that has already been put to practical use, in addition to those whose brewery characteristics are unknown because it is a newly produced raw material grain line. Also included are raw material grains that may have caused variations in brewing characteristics due to differences.

  Specific examples of the raw material cereals include rice, barley, barley malt, wheat, soybean and the like. Further, specific examples of the brew can include sake, shochu (especially rice shochu), beer, whiskey, soy sauce, miso and the like. Specific examples of combinations of brews and raw material grains include beer and malt (wheat malt); grains sprouted with whiskey; germinated grains; barley, wheat, soy sauce and wheat, soy; miso and wheat, rice, soy . In the following example, a prediction equation of the brewing characteristics of raw material rice in sake production is constructed, and the prediction equation can be used as it is to predict the brewing characteristics of raw material rice in rice shochu production, but the scope of the present invention is limited thereto. However, the method of the present invention can be applied to various raw material grains and brews.

  In the method of the present invention, a plurality of samples of brew material grains are used. From one raw grain sample, one or more samples for metabolomic analysis are prepared. Therefore, in the present invention, a plurality of metabolome analysis samples as many as or more than the number of raw grain samples are prepared. As for the raw material cereal samples, it is desirable to use a large number of samples having different varieties and cultivation environments as much as possible. When preparing only one sample for metabolomic analysis from one raw material grain sample and acquiring only one set of metabolomic analysis data, the minimum required number of raw material grain samples is 3 for verification equation construction, for verification There are four samples in total for one sample. That is, four or more samples for metabolome analysis are required. In order to obtain a prediction formula with sufficiently high prediction accuracy, it is desirable to prepare nine or more samples for metabolomic analysis (9 or more sets of metabolomic analysis data can be obtained). Nine or more samples of raw material cereal samples may be used, or three or more samples of raw material cereal samples may be prepared, and three or more samples for metabolomic analysis may be prepared from each of the raw material cereal samples.

  In metabolomic analysis, a plurality of metabolomic analysis samples are analyzed to obtain multiple sets of metabolomic analysis data. The metabolome analysis data is a data group obtained by comprehensively analyzing a group of compounds (mainly low molecular weight metabolites) contained in raw material grains using a sample for metabolome analysis prepared from raw material grains, It consists of quantitative numerical data on the abundance of each compound in the raw material grains. The quantitative numerical data as referred to herein is typically numerical data obtained by relatively quantifying the amount of each compound in the raw material grains.

  The compounds to be subjected to the metabolome analysis in the present invention include low molecular weight metabolites such as amino acids, amines, nucleotides, sugars and lipids, and DNA, as in the general meaning of "metabolome" in the field of metabolomics. Included are degradation products and fragments of high molecular weight biological compounds such as RNA, proteins and the like, and secondary metabolites. Therefore, in the present specification, a group of compounds to be analyzed in metabolome may be referred to as "metabolome". Although not particularly limited, in the present invention, metabolome analysis may be performed on a low molecular weight compound group having a molecular weight of about 50 to about 1,000.

  As a sample for metabolome analysis, an extract obtained by exhaustively extracting various compound groups contained in a raw material cereal sample may be used. For example, an extract obtained by crushing a raw material grain sample (raw rice such as raw rice grain) in a suitable solvent and extracting a compound group can be used as a metabolomic analysis sample. In the present invention, a hydrophilic compound group such as a lower alcohol solution can be preferably used as the solvent, since a group of hydrophilic compounds is mainly a target of metabolomic analysis. The lower alcohol solution may be, for example, an aqueous solution of methanol or alcohol having a concentration of about 30 to 70%. However, in the present invention, since a hydrophobic compound group may be used as a target of metabolomic analysis, it is also possible to use an extract obtained by extracting the compound group using a hydrophobic solvent as a sample for metabolomic analysis. The raw material grain extract may be subjected to ultrafiltration after removing the residue by centrifugation or the like after grain crushing.

  When the separation process of the compound in metabolome analysis is performed by gas chromatography, after extracting using an organic solvent such as diethyl ether, isopentane, pentane and the like, a sample derivatized so that the compound group to be analyzed is easily vaporized Or, the thermal decomposition product of the raw material grain sample can be used as a sample for metabolomic analysis. The above-mentioned method of derivatization of in vivo substances is a well-known conventional method.

  The analysis method itself in metabolomic analysis is not particularly limited as long as it is a method that can comprehensively analyze a group of compounds in a sample for metabolomic analysis. In general, metabolome analysis can be performed by liquid chromatography, capillary electrophoresis or a method combining gas chromatography and mass spectrometry or NMR analysis. Various liquid chromatographs, capillary electrophoresis, gas chromatographs, and mass spectrometry systems suitable for metabolome analysis are known, and there are also commercially available products. Metabolomic analysis can be performed using such commercially available analyzers. Alternatively, as metabolome analysis services are provided by various companies, such analysis services may be used to acquire a metabolome analysis data set.

  Liquid chromatography includes high performance liquid chromatography (HPLC), ultra high performance liquid chromatography (UHPLC, UPLC (registered trademark)). In metabolomic analysis for analyzing a large number of compounds, generally, high performance liquid chromatography is preferably used, but it is not particularly limited, and either HPLC or UHPLC may be used in the present invention. In general, hydrophilic interaction chromatography can be preferably used as the separation mode, but it is not limited to this, and the nature of the compound group extracted from the raw material rice, for example, is it hydrophilic or hydrophobic? Depending on the situation, liquid chromatography may be performed by combining an appropriate separation mode or mobile phase.

  In the case of employing gas chromatography, the gas chromatography step may be performed by a general method in the case of performing metabolomic analysis by GC / mass spectrometry.

  The method of mass spectrometry is also not particularly limited, and mass spectrometry generally employed in the metabolomics field and having a constant or higher quantitative performance can be used. As such mass spectrometry methods, a magnetic field type Fourier transform mass spectrometer, QMS (quadrupole mass spectrometry), TOF-MS (time-of-flight mass spectrometry), Triple QMS (triple quadrupole mass spectrometry) And QMS / TOF-MS etc., among which TOF-MS and QMS / TOF-MS can be preferably adopted, but not limited thereto. In the present invention, simply referring to “time-of-flight mass spectrometry” refers to mass analysis by a mass spectrometry system including a time-of-flight mass spectrometer, and includes TOF-MS and QMS / TOF-MS.

  Raw data on the content of each compound obtained by metabolomic analysis may be converted into numerical data of relative quantitative values using an appropriate software. As a method of calculating the relative quantitative value of each compound, the method of calculating as the relative area value of each peak of the mass spectrum, or the conical peak volume as the three-dimensional data of mass, time, and intensity of the measured value of the mass spectrum There is a method of calculating, etc., and any method may be adopted in the present invention. When using a commercially available analysis system for metabolomic analysis combining a chromatograph and a mass spectrometer, the system usually includes software for calculating relative quantitative values, so it is preferable to use such software. Just do it.

  In the present invention, the term "brewing property data" refers to various analysis data on raw material grains or straws and mash obtained therefrom, which are related to the brewing properties of raw material grains. Specific examples of brewing characteristics data when raw material grains are rice and brewed products are sake or shochu, analysis values of raw material rice obtained by sake-rice uniform analysis method, enzyme titer of rice bran produced from raw material rice The process of producing a sake or shochu from raw material rice may include the course of the course of moromi, and analysis values of general components and aroma components of sake or shochu produced from raw material rice. Analysis data including at least one, preferably a plurality of analysis values selected from these brewery-related analysis values may be acquired as brewery characteristic data. The persimmon, moromi and sake or shochu may be produced by a standard method for producing sake or shochu.

  Analysis value of raw material rice obtained by Sake-rice unified analysis method: Brown rice moisture of raw material rice, thousand grain weight before adjustment, thousand grain weight after adjustment, apparent grain proportion, true grain proportion, invalid grain proportion, broken grain rate, white rice moisture, 20 minutes water absorption It is an analysis value about rate, 120 minutes water absorption rate, steamed rice water absorption rate, digestibility brix, crude protein, and potassium. The analysis values of these 14 items can be obtained according to the conventional method according to the Rice National Standard Analysis Method for Sake Brewing Materials of Sake Rice Research Group.

  The enzyme titers of rice bran produced from raw material rice include α-amylase (AAase) activity, glucoamylase (GAase) activity, acid carboxypeptidase (ACPase) activity, acid protease (APase) activity, and α-glucosidase activity. Measurement values are included. It is also possible to produce rice bran by changing the proportion of rice used as raw material rice, and to measure the enzyme titer of each rice bran. The enzyme may be extracted from rice bran to make a rice bran enzyme solution according to the fourth revised National Tax Agency prescribed analysis method annotation, and various enzyme activities of this enzyme solution may be measured using a commercially available kit or the like.

  Analysis values of the course of fermentation (process of fermentation) in the process of producing sake or shochu from raw material rice, the maximum value of carbon dioxide weight loss per day, the number of days when the carbon dioxide weight loss per day reached the maximum value, carbon dioxide gas The integrated value of weight loss and the data of yield can be mentioned. The percolation rate is a value obtained by dividing the weight of the solid part when solid or liquid is separated by filtration, squeezing or the like of sorghum by centrifugation or filter paper, divided by the weight of raw rice and multiplied by 100. You can do a small charge test under general conditions and check these items. With regard to the moromi course, it is possible to carry out a small preparation test by changing the proportion of milled rice for raw rice, and to obtain analytical data of the moromi course from each test.

  As analytical values of the general components of the moromi supernatant or the sake liquor, there can be mentioned measured values of ethanol concentration, acidity, amino acidity, Japanese sake, and raw extract. Moreover, as an aroma component of the prepared sake, ethyl acetate (EtOAc), n-propyl alcohol (nPrOH), isobutyl alcohol (iBuOH), isoamyl acetate (iAmOAc), isoamyl alcohol (iAmOH), ethyl caproate (EtOCap) are mentioned. be able to. These components for producing sake may be measured in the usual manner. With regard to the general ingredients and aroma components of the made sake, conduct the small preparation test by changing the proportion of the raw rice rice as appropriate, and measure the general components and the aroma components of each of the obtained made liquor and use it as the brewing characteristics data. Can.

  As analytical values of general components of shochu, pH, acidity, ultraviolet absorption, thiobarbituric acid (TBA) value, coloring degree, 2,4,6-trichloroanisole (TCA) can be mentioned. As aroma components of shochu, acetaldehyde, ethyl acetate, n-propyl alcohol, isobutyl alcohol, isoamyl acetate, isoamyl alcohol, furfural, monoterpene alcohol (linalool, α-terpineol, citronellol, nerol, geraniol), ethyl capronate, Ethyl caprylate, ethyl caprate, β-phenethyl alcohol, β-phenethyl acetate, higher fatty acid ethyl ester (ethyl laurate, ethyl myristate, ethyl palmitate, ethyl linoleate, ethyl oleate, ethyl stearate) Can. These ingredients can also be used as brewing characteristics data by appropriately preparing shochu under general conditions by changing the proportion of rice used as the raw material rice, and measuring it according to a conventional method.

  Brewing property data are obtained for all of the raw grain samples to be subjected to metabolome analysis. The items of the brewing characteristics to be predicted may be selected, and the analysis of the raw material cereal samples may be performed on the items. In the following example, 28 × 3 samples for metabolomic analysis were prepared from a total of 28 samples of raw rice samples, and 28 × 3 metabolomic analysis data and 28 samples of brewing characteristics data were subjected to multiple regression analysis. 1000 grain weight before adjustment; 1000 grain weight after adjustment; true milled rice ratio; broken rice ratio; broken rice rate; 120 minutes water absorption rate; digestible brix; brown rice potassium content; α-amylase activity, glucoamylase activity, glucoamylase activity of rice bran prepared from raw material rice α-glucosidase activity, acid carboxypeptidase activity, and acid protease activity; Percentage ratio of moromi produced from raw material rice, maximum amount of CO 2 reduction per day, number of days when maximum amount of CO 2 reduction per day was reached, and carbon dioxide Integrated value of weight loss; ethanol content, acidity, amino acid degree, sake, ethyl acetate content, n-pe of sake produced from raw material rice It has been shown that a prediction equation with high prediction performance can be created for the ropyl alcohol content, isobutyl alcohol content, isoamyl acetate content, isoamyl alcohol content, and ethyl caproate content. The created prediction equation can be used as it is to predict the brewing characteristics of raw material rice in the production of shochu (especially rice shochu). Although these brewing characteristic items are specific examples of the items for which the prediction formula can be created by the method of the present invention, the scope of the present invention is not limited thereto.

  In the step of multiple regression analysis, multiple regression analysis is performed for each objective variable using two or more quantitative numerical data selected from metabolomic analysis data as an explanatory variable and brewing characteristic data as an objective variable to obtain a multiple regression equation . This multiple regression equation is a candidate for the prediction equation generated by the method of the present invention. The number of explanatory variables is not particularly limited, and for example, five or more, seven or more, or ten or more quantitative variable data may be used as the explanatory variables. Although it is possible to use all of the quantitative numerical data constituting one metabolome analysis data as an explanatory variable, it is general to use several to a dozen or so pieces of quantitative numerical data as an explanatory variable.

  As a method of multiple regression analysis, PLS (projections to latent structures) regression analysis or OPLS (orthogonal PLS) regression analysis can be preferably used. All of them are multivariate analysis methods widely known in the field of metabolomics, and many softwares therefor are commercially available, and such software can be used to perform these multiple regression analysis.

  Prior to the multiple regression analysis, standardization of explanatory variables (quantitative numerical data obtained by metabolomic analysis) and objective variables (brewing characteristic data) is performed by an appropriate standardization method. Usually, appropriate standardization tools are also included in software used for multiple regression analysis, so such tools may be used for standardization.

  In addition, metabolomic analysis data in the data set for multiple regression analysis may use all of the quantitative numerical data included in the data for multiple regression analysis, or selectively extract the quantitative numerical data based on a certain standard You may use for multiple regression analysis. For example, as is performed in the following embodiments, it is important to exclude large data from which quantitative numerical data are prominent or to extract quantitative numerical data with less variation among a plurality of metabolomic analysis data sets. It may be subjected to regression analysis. For each quantitative numerical data, for example, the ratio (standard deviation / average value × 100) of the standard deviation to the average value of the quantitative numerical data for each of a plurality of metabolomic analysis samples is calculated for each quantitative numerical data, It can be judged whether the ratio is low (for example, whether it is less than about 50%).

The metabolomic analysis data is subjected to multiple regression analysis using a partial set of a plurality of sets to create a prediction formula candidate, and then another metabolomic analysis data not used for generation of the prediction formula candidate. Applying a set of brewing characteristics data or another data set (a set of metabolomic analysis data and brewing characteristics data obtained separately using another cereal sample to be used for verification) to the prediction formula candidate, It is preferable to cross validate the prediction formula candidates and verify the prediction performance. In the verification of the prediction performance of the prediction formula candidate, for example, a prediction formula having a high R ² (determination coefficient) representing the degree of fitness or the degree of fitness can be evaluated as a prediction formula having a high prediction performance. Since R ² takes a value of 0 <R ² <1, for example, another data set not used for creating the prediction equation is applied to the prediction equation to obtain an R ² value, 0.7 ≦ R ² or 0.8 It is possible to evaluate that the prediction formula of ≦ R ² is high in prediction performance. In addition, the prediction performance of the prediction formula is the value of Root Mean Square Error of Prediction (RMSEP) (the sum of the squares of the differences between the individual predicted values and the actual values divided by the number of samples) or It is also possible to use the value of RPD (Residual Predictive Deviation) obtained by dividing the deviation by RMSEP. The RMSEP can evaluate a prediction formula having a small value as having high prediction performance. RPD can evaluate a prediction formula having a large value as having high prediction performance.

  When the prediction formula produced by the method of the present invention is applied to breeding of a new cereal variety, the prediction of the brewing characteristics using the prediction formula may be performed at any step of the breeding process. Selection may be performed by predicting the brewing characteristics at the individual selection step in the early stage of the breeding process, or selection by prediction of the brewing characteristics may be performed at the line selection step. By using the prediction formula of the present invention, it is possible to predict brewing characteristics using metabolomic analysis data obtained from a small amount of cereal samples, and therefore, it is possible to predict brewing characteristics even at an early stage of the breeding process where it is difficult to obtain a large amount of cereal samples. Selection can be carried out.

  Specifically, the method for producing cereal varieties using the brewing characteristics prediction equation can be carried out as follows.

  First, cereal samples are obtained from a growing cereal population or cereal lineage population. When the selection based on the brewing characteristics prediction is performed in the individual selection step, a grain sample is obtained from each individual of the growing grain group. In the case of performing selection based on brewing characteristics prediction in the step of line selection, at least one grain sample may be obtained from each line.

  Next, a sample for metabolome analysis is prepared from each grain sample, and metabolome analysis data is acquired. It is sufficient to acquire one set of metabolome analysis data from one cereal sample (ie, prepare one metabolome analysis sample from one cereal sample), but if necessary, acquire two or more sets of metabolome analysis data It is also good. Preparation of samples for metabolomic analysis and metabolomic analysis are carried out by the same method as adopted when formulating brewing characteristics prediction formula to be used.

  Then, the acquired metabolome analysis data is substituted into a brewing characteristic prediction equation to calculate a prediction value of the brewing characteristic. In this step, of course, only the necessary data (data about the compound (metabolome) employed in the prediction formula) may be used among the quantitative numerical data constituting the metabolome analysis data.

  Brewing characteristics are then evaluated based on the calculated predicted values, and individuals or lines predicted to have the desired brewing characteristics are selected from the above-mentioned cereal populations. Evaluation and selection based on predicted values may be performed according to the purpose of breeding. For example, in the case of predicting aroma components, when it is desired to breed a raw rice (rice) variety suitable for producing sake containing a large amount of ethyl caproate, it is assumed that the amount of ethyl caproate is large. It is sufficient to select a predicted individual or strain, and when it is desired to breed a raw material rice (rice) variety suitable for producing sake having a low ethyl caproate content, a selection is made of an individual or strain expected to have a low ethyl caproate production amount. do it.

  In breeding using a brewing characteristics prediction equation, a plurality of prediction equations related to a plurality of brewing characteristics can be used in combination. A prediction equation may be selected and used according to the purpose of breeding.

  Hereinafter, the present invention will be more specifically described based on examples. However, the present invention is not limited to the following examples.

1. Outline of prediction formula creation Predictive model with metabolomic analysis data as explanatory variable (x), various analysis data (sake-rice uniform analysis, enzyme concentration of persimmon, moromi course, general component analysis, aroma component analysis) as objective variable (y) Built. Data sets 1, 2 and 3 were created by creating data sets in which 1 set of various analysis data was combined with each of the 3 sets of metabolome analysis data. Furthermore, the data set integrated with Data set 1, 2, 3 is called Data set all. The model was constructed by orthogonal PLS (OPLS) regression analysis using SIMCA ver. 13.0.3 (Umetrics, Umea, Sweden). The outline is shown in FIG. In OPLS regression analysis, it is possible to create a prediction formula (for example, y = a * x1 + b * x2 ... + j * x10 + k) for finding the value of the objective variable (y) from the explanatory variable (x). Actual prediction formula creation should be performed independently with Data set 1, 2 and 3, and apply the created prediction formula to other data (Data set 2 and 3 when the prediction formula was created with Data set 1) Cross-validation and prediction performance verification.

  As brown rice samples, for each of five varieties including common sake rice varieties, multiple samples with different fields etc. were obtained, and a total of 28 samples were used.

2. Acquisition of metabolome analysis data (explanatory variable (x)) of brown rice extract (method)
For one brown rice sample, approximately 100 mg brown rice was precisely weighed and placed in a 2.0 mL screw cap microtube. To this tube was added a stainless steel ball of 5 mm in diameter (Inter RIKEN GAS MFG. Co., cat no. 1096-04). In addition, 10 times the volume of brown rice with 50% methanol (Milli-Q water and Methanol LC-MS CHROMASOLV® (Cat No. 34966-2.5 L) (Sigma-Aldrich Co. LLC, St. Louis, Mo., USA) Were mixed at a ratio of 1: 1 (v / v), and then allowed to stand at 4 ° C. for 30 minutes. Thereafter, the sample was pulverized using Micro Smash MS-100 (Tomy Seiko Co., Ltd., Tokyo, Japan) under the conditions of 300 ° C. and 3,500 rpm at 4 ° C. After grinding, the mixture was centrifuged at 15,000 rpm for 5 minutes at 4 ° C., and the supernatant was transferred to a 2.0 mL tube. This was centrifuged at 15,000 rpm for 10 minutes at 4 ° C., and the supernatant was subjected to ultrafiltration using an Amicon Ultra 3K-0.5 filter (Merck KGaA, Darmstadt, Germany) (4 ° C., 15,000 rpm, 60 minutes). The nominal molecular weight cut off of the filter used here is 3 kDa, and molecules of 3 kDa or more are considered to be removed from the sample.

  Three extractions were performed independently for each brown rice sample, and three samples for metabolomic analysis were prepared (a total of 28 × 3 samples of metabolome analysis samples were prepared for 28 samples of brown rice).

The 28 × 3 metabolomic analysis sample prepared above was subjected to UPLC-QTof MS, and metabolomic analysis was performed.
<Analytical conditions of UPLC-QTof MS>
The analysis was performed using ACQUITY UPLC system (Japan Waters Co., Ltd., Tokyo, Japan) and Waters Xevo QTOF MS (Japan Waters). The LC and MS conditions are shown below.
LC conditions: Column: ACQUITY UPLC HSS T3 Column, 2.1 × 150 mm 1.8 μm (Nippon Waters), column temperature: 40 ° C., flow rate: 0.3 mL / min, mobile phase: 0.1% formic acid (solution A) and 0.1% formic acid Acetonitrile (solution B), gradient; 0% solution B (0-5 min), 0-100% solution B (5-15 min), 100% solution B (15-20 min), 100-0% solution B ( 20-21 min), 0% solution B (21-30 min), sample injection volume: 3 μL
MS conditions: ionization; ESI positive ion mode, Desolvation gas flow; 800 L / h, Desolvation temperature; 450 ° C., Corn gas flow; 50 L / h, Source temperature; 140 ° C., Capillary voltage; 3 kV, Corn voltage; V, Collision energy; 6, MS mode; MS (m / z range 50-1,000), MS scan number; 0.2 sec

  From raw data obtained by analysis, using the supplied software MassLynxTM (Japan Waters) MarkerLynxTM XS (Japan Waters), the method of setting the error of accurate mass and retention time (Retention time) to 0.02 Da and 0.06 min, respectively The peak detection and alignment were performed using to create a marker table. As a result, 3,589 markers were obtained.

  Data from metabolomic analysis obtained from three samples for metabolomic analysis extracted from each brown rice sample at n = 3 are data 1, data 2 and data 3, respectively. Data from 28 brown rice samples are data 1, data 2 and data 3 Each set of 3 sets of metabolome analysis data was collected. These three sets of metabolomic analysis data were combined with various sets of analysis data 1 set of 28 brown rice samples obtained as described later, to obtain three multiple regression analysis data sets Data sets 1, 2, and 3. Data set 1, 2 and 3 integrated data set (ie, data set obtained by combining all 28 × 3 metabolome analysis data combined with various analysis data of 28 brown rice samples) as Data set all It was.

3. Acquisition of Various Analysis Data (Objective Variable (y)) The analysis items (objective variables) are shown in Tables 1-1 to 1-3.

  Hereinafter, analysis methods of the analysis items shown in Tables 1-1 to 1-3 will be described. For the sake-rice unified analysis, three analyzes were performed per brown rice sample, and the average value was taken as the analytical value of the brown rice sample. The other analysis values are prepared by adding rice bran with n = 2 to 1 brown rice sample, and measuring twice for each bran or preparation to obtain 4 measurement data, and the average value is used to obtain the relevant brown rice sample The analysis value of

<Sake rice uniform analysis>
14 items (Brown rice moisture, 1000 grain weight before adjustment, 1000 grain weight after adjustment, pure rice ratio, pure rice ratio, ineffective rice ratio, broken rice rate, broken rice rate, white rice water content, 20 minutes water absorption Analysis of rate, 120 minutes water absorption rate, steamed rice water absorption rate, digestibility brix, crude protein, potassium) was performed.

<Measurement of rice bran enzyme titer>
Enzyme titers (AAase, GAase, APase, ACPase, α-glucosidase) of rice bran prepared with 70% and 50% of rice were measured by the following method.

(Method)
Rice bran was produced in the following procedure. The brown rice was milled to 70% or 50% of the milling rate with a milling machine. Then, it was immersed in water and steamed. 60 mg of steamed rice was inoculated with conidia of 60 mg of Aspergillus oryzae RIBOIS 01 strain, and cultured at a temperature of 32 ° C. and a humidity of 90% for 18 hours. After the culture was mixed back, it was further cultured at a temperature of 32 ° C. and a humidity of 90% for 12 hours. The culture was again mixed and cultured at a temperature of 34 ° C. and a humidity of 80% for 3 hours. Thereafter, the cells were cultured at a temperature of 37 ° C. and a humidity of 70% for 3 hours, mixed with a culture, and cultured at a temperature of 40 ° C. and a humidity of 70% for 2 hours. Furthermore, the cells were cultured at a temperature of 42 ° C. and a humidity of 60% for 8 hours, and cultured for a total of 46 hours.

  To 5 g of rice bran, 25 mL of extraction buffer (0.5% NaCl, 10 mM acetate buffer) was added, and allowed to stand at 4 ° C. overnight. Thereafter, the solution was filtered using filter paper (ADVANTEC (registered trademark) No. 5A, Toyo Roshi Kaisha, Ltd., Tokyo, Japan), and the filtrate was used as a rice bran enzyme solution (the fourth revised National Tax Agency prescribed analysis method annotation). The prepared rice bran enzyme solution was measured for AAase, GAase, ACPase, APase and α-glucosidase activities by the following method.

α-Amylase (AAase):
It was measured using an α-amylase measurement kit (Kikkoman Biochemifa Co., Ltd., Tokyo, Japan). The rice bran enzyme solution was diluted 50 times with extraction buffer and measured according to the kit's measurement method.

Glucoamylase (GAase):
It measured using the saccharification force differential quantification kit (Kidkoman Biochemifa Inc.). The rice bran enzyme solution was diluted 2-fold with extraction buffer and measured according to the kit measurement method.

Acidic Carboxypeptidase (ACPase):
It measured using the acidic carboxypeptidase measurement kit (Kikkoman Biochemifa Inc.). The rice bran enzyme solution was diluted 5-fold with extraction buffer and measured according to the kit measurement method.

Acid Protease (APase):
The fourth revised National Tax Agency prescribed analysis method comment was followed. The rice bran enzyme solution was diluted 5-fold with extraction buffer and used for measurement.

α-glucosidase:
It measured using the saccharification force differential quantification kit (Kidkoman Biochemifa Inc.). The rice bran enzyme solution was used as it was, and it measured according to the measurement method of a kit.

<Small preparation test (fermentation progress)>
A small preparation test was conducted using 70% and 50% white rice, respectively.
・ Maximum value of CO2 weight loss per day ・ Number of days when CO2 weight loss amounted to the maximum value per day ・ Integrated value of CO2 weight loss amount ・ Percentage ratio was examined.

(Method)
Sake yeast K1801 strain was precultured by standing overnight at 30 ° C. in 5 mL of Koji extract medium (BOME 10). 200 μL of the preculture liquid was transferred to 200 mL of anther extract medium, and main culture was performed by standing at 30 ° C. for two nights. The main culture was centrifuged at 4 140 g at 4 ° C. for 10 minutes to recover yeast cells. The yeast cells were adjusted to 1.0 × 10 ⁸ cells / mL and used as a yeast suspension.

  Then, preparation was performed in the following procedure. Add rice bran equivalent to 20 g of white rice (rice bran prepared from 20 g white rice) to a 450 mL mayonnaise bottle containing 129 g (mL) of sterile water and 90 μL of 50% lactic acid, and leave it at 4 ° C for 1 hour did. Then, 1 mL of yeast suspension and 80 g of white rice (70% or 50% of polished rice) were steamed to prepare hanging rice, which was mixed well. Moromi were incubated at 15 ° C. for 19 days while weighing daily. Thereafter, the moromi was centrifuged at 4 140 g at 4 ° C. for 15 minutes, and the supernatant was recovered. Thereafter, the supernatant was used for analysis as produced sake. The ratio of the precipitation weight to the raw material white rice was calculated as the ratio of rice bran.

<Analysis of sake making>
The produced sake obtained by the above-mentioned small preparation test was analyzed, and the following items were examined.
70% and 50% ethanol concentration, ethanol concentration, acidity / amino acid degree, Japanese sake, aroma components (EtOAc, nPrOH, iBuOH, iAmOAc, iAmOH, EtOCap), raw extract

(Method)
Ethanol concentration:
It quantified by the gas chromatography using AOC-20iAUTO INJECTOR (Shimadzu Corporation, Kyoto, Japan) and GC-17A GAS CHROMATOGRAPH (Shimadzu Corporation). 40 μL samples were taken in 1.5 mL vials and mixed with 960 μL 1% isopropanol (internal standard). The alcohol was separated on a DB-624 capillary column (id 0.53 mm, length 30 m, film thickness 3 μm; Agilent Technologies Inc., Santa Clara, Calif., USA). The conditions are as follows. Injection temperature, 250 ° C .; oven temperature, 50 ° C .; detector temperature, 250 ° C., carrier gas, He 6.0 mL / min. (Column inlet pressure 28 kPa, split ratio 1:40)

Acidity / Amino acid:
The acidity was measured in accordance with the 4th revised National Tax Agency's prescribed analysis method annotation (pp. 20-23), and the amino acidity was measured according to "Study on the method for analyzing amino acidity of sake using ethanol".

Japanese sake degree:
For Japanese sake, specific gravity S (15/4 ° C) (specific gravity to the density of water at 4 ° C of 15 ° C) Density / Specific Gravity Meter DA-520 (KYOTO ELECTRONICS MANUFACTURING CO., LTD, Kyoto, Japan ) And was calculated at 1443 / S-1443.

Aroma component:
Volatile aromatic compounds were measured using the 7697A Headspace Sampler (Agilent Technologies) and the 7890B GC system (Agilent Technologies) in the same manner as T. Katou et al. Yeast 2008; 25: 799-807. .

Original extract:
Theoretical extract (mostly glucose). From the extract and alcohol in persimmon and sake, it is calculated that the extract necessary to ferment and produce them is present in persimmon and sake before fermentation. Calculated using sake manufacturing technology (p. 247) and the 4th Revised National Tax Agency prescribed analysis method annotation (pp. 24-25) as a reference.

4. Statistical analysis (construction of prediction formula)
The actual work procedure for creating a prediction equation is shown in FIG.

  First, OPLS regression analysis is performed using SIMCA ver. 13.0.3 (Umetrics, Umea, Sweden) with brown rice metabolome analysis data as explanatory variables (x) (the following annotations *) and various analysis data as target variables (y). The Regression analysis was performed independently for each Data set or all Data sets of brown rice metabolome analysis data. Prior to analysis, standardization was performed using Pareto scaling ((Xi-mean) / √SD)) as the explanatory variable and Autoscaling ((Xi-mean) / SD) as the target variable. Autoscaling is a normalization that makes the mean 0 and the variance 1 and Pareto scaling is a standardization that leaves a large value than Autoscaling.

Next, we narrowed down the explanatory variables in order to create a practical and highly accurate prediction formula. The model was analyzed by S-plot, and a total of 10 explanatory variables were selected, each having a large y-axis absolute value common to each data set. Similarly, OPLS regression analysis was performed for each data set using latent variables as 1 + 4 + 0 using only these 10 explanatory variables. It was confirmed that this latent variable was not overfit by Permutation test. Next, another Data set or Data set all was applied to each prediction formula, and a prediction formula having a high R ² value was evaluated as an optimum prediction formula.

(Annotation *) The explanatory variable brown rice metabolome analysis data is composed of the intensities of 3,589 markers. Among the 3,585 pieces excluding the four data including intensity and 10 ²² too and all data (ALL), it was used for statistical analysis. Furthermore, for each marker, calculate the ratio (stdev / average * 100) of the standard deviation to the average value for each raw rice sample (3 sets), and the 751 markers that are less than 50% have less variance (< 50%) and used for statistical analysis.

5. Evaluation Results of Prediction Formulas Table 2-1 to Table 2-3 show values of R ² (degree of fitness) and Q ² (predictability) of each prediction formula created above. Tables 3-1 to 3-9 show the average values of R ² values when each Data set (1, 2, 3, all) is applied to each prediction equation, and the R ² values of each prediction equation. is there. For the explanatory variables shown in Tables 3-1 to 3-9, a total of six prediction equations (1 to 1, excluding all) created with all data (ALL) and data with less variation (<50%) 3) We selected better and better models. Among the six prediction formulas, those with an R ² average value of 0.7 ≦ R ² Avg. <0.8 after application were selected as good models, and those with 0.8 ≦ R ² Avg. The actual prediction formulas of the selected model are shown in Tables 4-1 to 4-4.

  The above results are summarized in Table 5. Of the 54 objective variables, 25 were able to construct an excellent model ("excellent" in the table). Of these, 14 models were all data (All), and 11 models were selected from data with less variation (<50%). Similarly, a good model could be constructed with 11 explanatory variables ("Good" in the table). Of these, 6 models were all data (All), and 5 models with less variation (<50%) were selected. From this, it can not be said that data with less variation can always create a good prediction formula.

  From the above, it was shown that it is possible to construct a model that predicts the characteristics of raw rice, the enzyme titer of malt, the moromi course, and the characteristics of mature sake with 36 out of 54 objective variables. The prediction equation obtained here can be used as it is not only for sake but also as a prediction equation for rice shochu (a prediction equation for brewing characteristics of raw material rice of rice shochu).

Claims (12)

A method of creating a brewing characteristic prediction equation for predicting brewing characteristics when producing brews using raw material cereals for raw material cereals of which the brewing characteristics are unknown,
Preparing one or more samples for metabolomic analysis from each of a plurality of brew raw material cereal samples;
Using the metabolome analysis sample, metabolomic analysis is conducted to comprehensively analyze compounds contained in raw material grains, and metabolome analysis data composed of quantitative numerical data on the abundance of each compound in raw material grains is used. Acquiring process;
Obtaining at least one brewing characteristic data related to brewing characteristics for each of said brew material cereal samples;
Performing multiple regression analysis for each objective variable using two or more quantitative numerical data selected from each metabolome analysis data as an explanatory variable, and using a brewery characteristic data as an objective variable;
Method, including.   The method according to claim 1, wherein the raw material cereal is rice, barley, barley malt, wheat or soybean.   The method according to claim 1 or 2, wherein the brew is sake, shochu, beer, whiskey, soy sauce, or miso.   The method according to claim 1, wherein the raw material cereal is rice and the brewed product is sake or shochu.   The above-mentioned brewery characteristic data is an analysis value of raw material rice obtained by sake-rice unified analysis method, an enzyme titer of rice bran produced from raw material rice, a course of mash in the process of producing sake or shochu from raw material rice, and raw material rice The method according to claim 4, comprising at least one selected from analysis values of general components and aroma components of manufactured sake or shochu.   Perform multiple regression analysis using a part of multiple metabolome analysis data, create a prediction formula candidate, and then cross-validate the other prediction formula candidate with another metabolome analysis data, and perform the prediction formula 6. A method according to any one of the preceding claims, comprising verifying the predictive performance of the candidate.   The method according to any one of claims 1 to 6, wherein the multiple regression analysis is OPLS regression analysis.   The method according to any one of claims 1 to 7, wherein the sample for metabolomic analysis is an extract obtained by grinding raw material grains in a hydrophilic solvent to extract a group of compounds.   The method according to any one of claims 1 to 8, wherein the metabolome analysis is performed by liquid chromatography or a combination of gas chromatography and mass spectrometry.   10. The method of claim 9, wherein the mass spectrometry is time of flight mass spectrometry.   The method according to any one of claims 1 to 10, wherein five or more quantitative numerical data are used as an explanatory variable. A step of creating a brewing characteristic prediction equation by the method according to any one of claims 1 to 11;
Obtaining a cereal sample from a growing population of cereals or a population of cereal lines, preparing a sample for metabolomic analysis from the sample, and acquiring metabolomic analysis data;
Substituting the acquired metabolomic analysis data into a brewing characteristic prediction equation to calculate a prediction value of the brewing characteristic;
Selecting from the population individuals or lines predicted to have desired brewing characteristics based on the predicted value;
How to produce cereal varieties, including