阅读说明：本技术 一种快速识别大曲等级的方法 (Method for rapidly identifying Daqu grade ) 是由 王寿峰 明红梅 周慧 周冯利 于 2020-06-01 设计创作，主要内容包括：本发明涉及大曲识别技术领域,具体涉及一种快速识别大曲等级的方法。具体技术方案为：一种快速识别大曲等级的方法,采集多个不同等级的大曲样品的核磁图谱,并对核磁图谱进行校正,分段积分；然后构建大曲等级识别分类模型；最后将未知大曲的核磁数据输入分类模型中,实现对大曲等级的快速识别。本发明运用<Sup>1</Sup>H-NMR的技术,建立了<Sup>1</Sup>H-NMR积分数据库,构建大曲等级识别分类模型,最终与计算机联用对大曲等级进行识别。(The invention relates to the technical field of yeast identification, in particular to a method for quickly identifying the grade of yeast. The specific technical scheme is as follows: a method for rapidly identifying the grade of Daqu comprises the steps of collecting nuclear magnetic maps of Daqu samples with different grades, correcting the nuclear magnetic maps, and performing segmentation integration; then constructing a yeast grade identification classification model; and finally, inputting the nuclear magnetic data of the unknown Daqu into a classification model to realize the rapid identification of the Daqu grade. Application of the invention 1 H-NMR technique, established 1 And (4) constructing a Daqu grade identification classification model by using an H-NMR integral database, and finally identifying the Daqu grade by using the Daqu grade identification classification model and a computer.)

1. A method for rapidly identifying the grade of a yeast is characterized in that: collecting nuclear magnetic maps of a plurality of Daqu samples with different grades, correcting the nuclear magnetic maps, and performing segmentation integration; then constructing a yeast grade identification classification model; and finally, inputting the nuclear magnetic data of the unknown Daqu into a classification model to realize the rapid identification of the Daqu grade.

2. The method for rapidly identifying the level of the yeast according to claim 1, wherein: the preparation process of the Daqu sample comprises the following steps: adding distilled water into the Daqu sample, grinding, centrifuging, taking supernatant, adding into a nuclear magnetic tube, adding heavy water, mixing, sealing, refrigerating, and testing.

3. The method for rapidly identifying the level of the yeast according to claim 2, wherein: the grinding process comprises the following steps: grinding at 70Hz for 300s, and circulating for 2 times; the centrifugation process is as follows: centrifuging at 10000rpm for 10 min.

4. The method for rapidly identifying the level of the yeast according to claim 2, wherein: deuterium-depleted trimethylsilylpropionic acid sodium was contained in the heavy water in an amount of 0.05%.

5. The method for rapidly identifying the level of the yeast according to claim 1, wherein: and the nuclear magnetic spectrum is processed by Mestrenova software, corrected and integrated in sections, and a standard sample nuclear magnetic integration database is established.

6. The method for rapidly identifying the grade of the yeast according to claim 5, wherein: the method for constructing the Daqu grade recognition classification model is a support vector machine or a neural network.

7. The method for rapidly identifying the level of the yeast according to claim 6, wherein: analyzing data in a standard sample nuclear magnetic integral database through a support vector machine, and establishing a classification model; and then introducing the nuclear magnetic spectrum data of the unknown Daqu sample into a classification model, and identifying the unknown Daqu sample.

8. The method as claimed in claim 6, wherein the data in the NMR database of the standard sample is analyzed by neural network, the number of the network layers is 3, the number of the hidden nodes in the first layer is 20, and the maximum number of the iterations is 5 × 10⁴Training and optimizing to establish a classification model; and then introducing the nuclear magnetic spectrum data of the unknown Daqu sample into a classification model, and identifying the unknown Daqu sample.

Technical Field

The invention relates to the technical field of yeast identification, in particular to a method for quickly identifying the grade of yeast.

Background

The Daqu is a microecological product which is made by fermenting wheat mainly or supplemented with peas, barley and the like, has a large shape and contains various fungi, enzymes and fragrant substances, is a saccharification leavening agent and a fragrance-producing agent in wine production, and is also a partial raw material in the wine production. The Daqu contains rich flavor substances and precursors thereof due to the metabolism of microorganisms, and the flavor compounds form the 'Daqu incense' of the Daqu and have great influence on the production of Chinese liquor.

The quality evaluation of the Daqu mostly takes a traditional sensory evaluation mode as a main part and a rational analysis as an auxiliary part, although the quality of the Daqu can be reflected to a certain extent, some scientific bases are lacked, physicochemical indexes are constant analysis, and individual indexes can be completed within 7 days, so that the method is time-consuming and labor-consuming (a general analysis method for brewing Daqu QB/T4257-2011; CN104865360A, a method and a system for evaluating the quality of the strong-flavor Daqu).

In recent years, with the increasing of analytical instruments and analytical techniques, great progress has been made in analyzing the flavor substance composition of the Daqu, and reasonable analysis of the Daqu flavor substance can reflect the quality of the Daqu to a certain extent. Chinese patent (CN103293264A, a method for identifying the quality of Daqu) introduces a method which can distinguish the finished product koji from the white koji and the discharged koji by applying solid phase microextraction and combining with a gas chromatography-mass spectrometry technology, and has low accuracy. The high-throughput DNA sequencing technology rapidly developed in recent years can obtain the composition of a microbial community, and the technology is applied to the research of Daqu microorganisms, so that the relationship between the microbial community structure and the Daqu quality can be disclosed (Chinese patent CN104109719A, a method for identifying the Daqu quality based on a box diagram, CN104372075A, a method for identifying the Daqu quality by constructing a discrimination model, and CN109949863A, a method for identifying the Daqu quality based on a random forest model).

Disclosure of Invention

Aiming at the defects of the prior art, the invention provides a method for quickly identifying the grade of the yeast for making hard liquor, and the application¹H-NMR technique, established¹And (4) constructing a Daqu grade identification classification model by using an H-NMR integral database, and finally identifying the Daqu grade by using the Daqu grade identification classification model and a computer.

In order to achieve the purpose, the invention is realized by the following technical scheme:

the invention discloses a method for rapidly identifying the grade of yeast for making hard liquor, which comprises the steps of collecting nuclear magnetic maps of a plurality of yeast for making hard liquor with different grades, correcting the nuclear magnetic maps and performing segmentation integration; then constructing a yeast grade identification classification model; and finally, inputting the nuclear magnetic data of the unknown Daqu into a classification model to realize the rapid identification of the Daqu grade.

Preferably, the preparation process of the Daqu sample comprises the following steps: adding distilled water into the Daqu sample, grinding, centrifuging, taking supernatant, adding into a nuclear magnetic tube, adding heavy water, mixing, sealing, refrigerating, and testing.

Preferably, the grinding process is as follows: grinding at 70Hz for 300s, and circulating for 2 times; the centrifugation process is as follows: centrifuging at 10000rpm for 10 min.

Preferably, the deuterium depleted water contains 0.05% of deuterated trimethylsilyl sodium propionate.

Preferably, the nuclear magnetic spectrum is processed by Mestrenova software, and the nuclear magnetic spectrum is corrected and integrated in a segmented manner to establish a standard sample nuclear magnetic integration database.

Preferably, the method for constructing the Daqu grade recognition classification model is a support vector machine or a neural network.

Preferably, analyzing data in a standard sample nuclear magnetic integral database through a support vector machine, and establishing a classification model; and then introducing the nuclear magnetic spectrum data of the unknown Daqu sample into a classification model, and identifying the unknown Daqu sample.

Preferably, the data in the nuclear magnetic integration database of the standard sample is subjected to neural network analysis, the number of network layers is 3, the number of hidden nodes in the first layer is 20, and the maximum iteration number is 5 × 10⁴Training and optimizing to establish a classification model; and then introducing the nuclear magnetic spectrum data of the unknown Daqu sample into a classification model, and identifying the unknown Daqu sample.

The invention has the following beneficial effects:

1) compared with the prior art, the invention adopts high-flux and high-resolution¹And (3) evaluating the quality of the Daqu of different varieties by using an H-NMR metabonomics technology. The method shows that the compositions and the contents of the metabolites of the yeast of different varieties have obvious difference, and nuclear magnetic spectrum is corrected and segmented and integrated; constructing the grade of the yeast by adopting a data processing technologyAnd identifying a classification model to realize the rapid identification of the unknown yeast grade. The invention can be used for establishing nuclear magnetic resonance integral data databases of Daqu with different qualities. The method has the characteristics of systematic, accurate, reliable and efficient, and has important significance for perfecting the current Daqu quality evaluation technology system by combining with the computer statistical analysis technology.

2) The invention fundamentally solves the problems brought by artificial sensory evaluation and main chemical component analysis: the artificial sensory evaluation workload is large, the result is limited by a plurality of environmental factors, human psychological factors and the like, the subjective performance and the randomness are very large, and the sensory evaluation and the physical and chemical analysis are combined to be a basic method for evaluating the quality of the Daqu, but the physical and chemical indexes are limited, the components of the Daqu are very complex, certain simple indexes are not enough to reflect the inherent quality of the Daqu, and the application has serious limitation.

3) The quality of the yeast for making hard liquor provided by the invention¹The H-NMR evaluation method is a good means for analyzing a sample with complex components, the integrity and the fuzziness of the H-NMR evaluation method can reflect the overall information of the sample, and the workload is not large. More importantly, the method can achieve the purpose of distinguishing different types, different regions and different qualities of the yeast.

Drawings

FIG. 1 is a graph of the magnetohydrogen spectrum of Daqu nucleus;

FIG. 2 is a graph of different grades of Daqu O2PLS-DA analysis;

FIG. 3 is an analysis chart of unknown grade Daqu O2 PLS-DA.

Detailed Description

The technical solutions in the embodiments of the present invention will be clearly and completely described below with reference to the drawings in the embodiments of the present invention, and it is obvious that the described embodiments are only a part of the embodiments of the present invention, and not all of the embodiments. All other embodiments, which can be derived by a person skilled in the art from the embodiments given herein without making any creative effort, shall fall within the protection scope of the present invention.

Unless otherwise indicated, the technical means used in the examples are conventional means well known to those skilled in the art.

The invention discloses a method for rapidly identifying the grade of yeast for making hard liquor, which comprises the following steps of collecting nuclear magnetic maps of a plurality of yeast for making hard liquor with different grades, correcting the nuclear magnetic maps, and performing segmentation integration, wherein the method specifically comprises the following steps: and (3) processing the nuclear magnetic spectrum through Mestrenova software, correcting and performing segmented integration on the nuclear magnetic spectrum, and establishing a standard sample nuclear magnetic integration database. Then, constructing a Daqu grade recognition classification model by any one of a partial least square method discriminant analysis method, a support vector machine and a neural network; and finally, inputting the nuclear magnetic data of the unknown Daqu into a classification model to realize the rapid identification of the Daqu grade.

The preparation process of the Daqu sample comprises the following steps: adding distilled water into Daqu sample, grinding, centrifuging, adding supernatant into nuclear magnetic tube, adding deuterium-enriched water containing 0.05% deuterated trimethylsilyl sodium propionate (TMSP) internal standard, mixing, sealing, refrigerating, testing,¹the measurement conditions for H-NMR were: the measuring frequency is 600.13MHz, the temperature is 298K, the scanning range is-5-15 ppm, the sampling point is 64K, and the scanning frequency is 64. The grinding process comprises the following steps: grinding at 70Hz for 300s, and circulating for 2 times; the centrifugation process is as follows: centrifuging at 10000rpm for 10 min.

When the partial least square method is adopted for discriminant analysis: analyzing the data in the standard sample nuclear magnetic integral database by using partial least squares discriminant analysis O2PLS-DA to establish a classification model; and then introducing the nuclear magnetic spectrum data of the Daqu sample at the position into a classification model, and identifying the unknown Daqu sample.

When the support vector machine is adopted for analysis: analyzing data in a standard sample nuclear magnetic integral database through a support vector machine, and establishing a classification model; and then introducing the nuclear magnetic spectrum data of the unknown Daqu sample into a classification model, and identifying the unknown Daqu sample.

When the neural network analysis is adopted, the data in the standard sample nuclear magnetic integral database is subjected to the neural network analysis, the number of network layers is 3, the number of hidden nodes in the first layer is 20, and the maximum iteration number is 5 × 10⁴Training and optimizing to establish a classification model; then introducing the nuclear magnetic spectrum data of the unknown Daqu sample into a classification model, and carrying out classification on the unknown Daqu sampleAnd identifying the Daqu sample.

The invention is further illustrated below with reference to specific examples.