CN109253983B - Method for rapidly identifying and detecting parvalbumin based on mid-infrared spectrum and neural network technology - Google Patents
Method for rapidly identifying and detecting parvalbumin based on mid-infrared spectrum and neural network technology Download PDFInfo
- Publication number
- CN109253983B CN109253983B CN201811455680.0A CN201811455680A CN109253983B CN 109253983 B CN109253983 B CN 109253983B CN 201811455680 A CN201811455680 A CN 201811455680A CN 109253983 B CN109253983 B CN 109253983B
- Authority
- CN
- China
- Prior art keywords
- parvalbumin
- neural network
- infrared
- mid
- original
- Prior art date
- Legal status (The legal status is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the status listed.)
- Active
Links
- 102000001675 Parvalbumin Human genes 0.000 title claims abstract description 66
- 108060005874 Parvalbumin Proteins 0.000 title claims abstract description 66
- 238000000034 method Methods 0.000 title claims abstract description 38
- 238000002329 infrared spectrum Methods 0.000 title claims abstract description 31
- 238000013528 artificial neural network Methods 0.000 title claims abstract description 26
- 238000005516 engineering process Methods 0.000 title claims abstract description 23
- 241000251468 Actinopterygii Species 0.000 claims abstract description 25
- 108090000623 proteins and genes Proteins 0.000 claims abstract description 22
- 102000004169 proteins and genes Human genes 0.000 claims abstract description 21
- 238000001514 detection method Methods 0.000 claims abstract description 19
- 239000013598 vector Substances 0.000 claims abstract description 17
- 238000003062 neural network model Methods 0.000 claims abstract description 16
- 238000004476 mid-IR spectroscopy Methods 0.000 claims abstract description 13
- 238000010521 absorption reaction Methods 0.000 claims abstract description 11
- 238000005259 measurement Methods 0.000 claims abstract description 9
- 238000012360 testing method Methods 0.000 claims abstract description 7
- 238000009795 derivation Methods 0.000 claims abstract description 6
- 238000001228 spectrum Methods 0.000 claims description 27
- 230000003595 spectral effect Effects 0.000 claims description 8
- 241000252234 Hypophthalmichthys nobilis Species 0.000 claims description 7
- 238000004566 IR spectroscopy Methods 0.000 claims description 7
- 238000012549 training Methods 0.000 claims description 6
- 238000004458 analytical method Methods 0.000 claims description 5
- 241001275898 Mylopharyngodon piceus Species 0.000 claims description 4
- 241000404975 Synchiropus splendidus Species 0.000 claims description 4
- 241000376029 Tachysurus fulvidraco Species 0.000 claims description 4
- 241001441722 Takifugu rubripes Species 0.000 claims description 4
- 241000276707 Tilapia Species 0.000 claims description 4
- 238000004364 calculation method Methods 0.000 claims description 4
- 241001233037 catfish Species 0.000 claims description 4
- 238000012937 correction Methods 0.000 claims description 4
- 238000009499 grossing Methods 0.000 claims description 4
- 238000010606 normalization Methods 0.000 claims description 4
- 241001609213 Carassius carassius Species 0.000 claims description 3
- 241001597062 Channa argus Species 0.000 claims description 3
- 241000252230 Ctenopharyngodon idella Species 0.000 claims description 3
- 238000007781 pre-processing Methods 0.000 claims description 2
- 238000013526 transfer learning Methods 0.000 claims description 2
- 241001441723 Takifugu Species 0.000 claims 1
- 239000013566 allergen Substances 0.000 abstract description 9
- 239000003153 chemical reaction reagent Substances 0.000 abstract description 2
- 239000000047 product Substances 0.000 description 16
- 239000006228 supernatant Substances 0.000 description 10
- 239000013505 freshwater Substances 0.000 description 9
- 239000002244 precipitate Substances 0.000 description 6
- 238000007637 random forest analysis Methods 0.000 description 6
- 238000012706 support-vector machine Methods 0.000 description 6
- 238000002965 ELISA Methods 0.000 description 4
- 238000002156 mixing Methods 0.000 description 4
- 238000003756 stirring Methods 0.000 description 4
- 241000054450 Takifugu flavidus Species 0.000 description 3
- YBYRMVIVWMBXKQ-UHFFFAOYSA-N phenylmethanesulfonyl fluoride Chemical group FS(=O)(=O)CC1=CC=CC=C1 YBYRMVIVWMBXKQ-UHFFFAOYSA-N 0.000 description 3
- 238000003752 polymerase chain reaction Methods 0.000 description 3
- 239000000843 powder Substances 0.000 description 3
- 241000252073 Anguilliformes Species 0.000 description 2
- 108010028690 Fish Proteins Proteins 0.000 description 2
- 229940124158 Protease/peptidase inhibitor Drugs 0.000 description 2
- 150000001408 amides Chemical class 0.000 description 2
- 239000001110 calcium chloride Substances 0.000 description 2
- 229910001628 calcium chloride Inorganic materials 0.000 description 2
- 238000004422 calculation algorithm Methods 0.000 description 2
- 230000000052 comparative effect Effects 0.000 description 2
- 230000007547 defect Effects 0.000 description 2
- 238000001962 electrophoresis Methods 0.000 description 2
- 230000001900 immune effect Effects 0.000 description 2
- 239000012535 impurity Substances 0.000 description 2
- 239000000203 mixture Substances 0.000 description 2
- 239000000137 peptide hydrolase inhibitor Substances 0.000 description 2
- 238000012545 processing Methods 0.000 description 2
- YNJBWRMUSHSURL-UHFFFAOYSA-N trichloroacetic acid Chemical compound OC(=O)C(Cl)(Cl)Cl YNJBWRMUSHSURL-UHFFFAOYSA-N 0.000 description 2
- VZSRBBMJRBPUNF-UHFFFAOYSA-N 2-(2,3-dihydro-1H-inden-2-ylamino)-N-[3-oxo-3-(2,4,6,7-tetrahydrotriazolo[4,5-c]pyridin-5-yl)propyl]pyrimidine-5-carboxamide Chemical compound C1C(CC2=CC=CC=C12)NC1=NC=C(C=N1)C(=O)NCCC(N1CC2=C(CC1)NN=N2)=O VZSRBBMJRBPUNF-UHFFFAOYSA-N 0.000 description 1
- QKNYBSVHEMOAJP-UHFFFAOYSA-N 2-amino-2-(hydroxymethyl)propane-1,3-diol;hydron;chloride Chemical compound Cl.OCC(N)(CO)CO QKNYBSVHEMOAJP-UHFFFAOYSA-N 0.000 description 1
- 241000252211 Carassius Species 0.000 description 1
- 241001417978 Channidae Species 0.000 description 1
- 241000252228 Ctenopharyngodon Species 0.000 description 1
- 208000004262 Food Hypersensitivity Diseases 0.000 description 1
- 206010016946 Food allergy Diseases 0.000 description 1
- 238000005033 Fourier transform infrared spectroscopy Methods 0.000 description 1
- 108010033276 Peptide Fragments Proteins 0.000 description 1
- 102000007079 Peptide Fragments Human genes 0.000 description 1
- 241000252496 Siluriformes Species 0.000 description 1
- 241000269319 Squalius cephalus Species 0.000 description 1
- 238000000862 absorption spectrum Methods 0.000 description 1
- 230000000172 allergic effect Effects 0.000 description 1
- 125000003275 alpha amino acid group Chemical group 0.000 description 1
- 230000003321 amplification Effects 0.000 description 1
- 208000010668 atopic eczema Diseases 0.000 description 1
- 230000009286 beneficial effect Effects 0.000 description 1
- 238000009739 binding Methods 0.000 description 1
- 239000007853 buffer solution Substances 0.000 description 1
- 230000002860 competitive effect Effects 0.000 description 1
- 238000010411 cooking Methods 0.000 description 1
- 238000000605 extraction Methods 0.000 description 1
- 235000013305 food Nutrition 0.000 description 1
- 235000020932 food allergy Nutrition 0.000 description 1
- 239000012634 fragment Substances 0.000 description 1
- 238000004108 freeze drying Methods 0.000 description 1
- 238000001294 liquid chromatography-tandem mass spectrometry Methods 0.000 description 1
- 238000004949 mass spectrometry Methods 0.000 description 1
- 238000001819 mass spectrum Methods 0.000 description 1
- 210000003205 muscle Anatomy 0.000 description 1
- 238000003199 nucleic acid amplification method Methods 0.000 description 1
- 238000001556 precipitation Methods 0.000 description 1
- 238000002360 preparation method Methods 0.000 description 1
- 239000002994 raw material Substances 0.000 description 1
- 238000003753 real-time PCR Methods 0.000 description 1
- 235000021067 refined food Nutrition 0.000 description 1
- 238000012216 screening Methods 0.000 description 1
- 239000007787 solid Substances 0.000 description 1
- 239000000243 solution Substances 0.000 description 1
- 239000000126 substance Substances 0.000 description 1
- 238000002198 surface plasmon resonance spectroscopy Methods 0.000 description 1
- 208000024891 symptom Diseases 0.000 description 1
- 238000012795 verification Methods 0.000 description 1
Images
Classifications
-
- G—PHYSICS
- G01—MEASURING; TESTING
- G01N—INVESTIGATING OR ANALYSING MATERIALS BY DETERMINING THEIR CHEMICAL OR PHYSICAL PROPERTIES
- G01N21/00—Investigating or analysing materials by the use of optical means, i.e. using sub-millimetre waves, infrared, visible or ultraviolet light
- G01N21/17—Systems in which incident light is modified in accordance with the properties of the material investigated
- G01N21/25—Colour; Spectral properties, i.e. comparison of effect of material on the light at two or more different wavelengths or wavelength bands
- G01N21/31—Investigating relative effect of material at wavelengths characteristic of specific elements or molecules, e.g. atomic absorption spectrometry
- G01N21/35—Investigating relative effect of material at wavelengths characteristic of specific elements or molecules, e.g. atomic absorption spectrometry using infrared light
- G01N21/3563—Investigating relative effect of material at wavelengths characteristic of specific elements or molecules, e.g. atomic absorption spectrometry using infrared light for analysing solids; Preparation of samples therefor
Landscapes
- Physics & Mathematics (AREA)
- Spectroscopy & Molecular Physics (AREA)
- Health & Medical Sciences (AREA)
- Life Sciences & Earth Sciences (AREA)
- Chemical & Material Sciences (AREA)
- Analytical Chemistry (AREA)
- Biochemistry (AREA)
- General Health & Medical Sciences (AREA)
- General Physics & Mathematics (AREA)
- Immunology (AREA)
- Pathology (AREA)
- Investigating Or Analysing Materials By Optical Means (AREA)
Abstract
The invention belongs to the technical field of biology, and provides a method for rapidly identifying and detecting major allergen parvalbumin in fish aquatic products based on mid-infrared spectroscopy and neural network technology. The method comprises the following steps: (1) performing infrared spectrum measurement on protein extracted from a sample to be measured to obtain original intermediate infrared spectrogram data; (2) obtaining a second derivative map by the original intermediate infrared spectrogram through second derivation; (3) acquiring three-dimensional vector data from an original infrared spectrum and a second derivative spectrogram thereof, and identifying by using a trained neural network model to obtain a classification result: the three-dimensional vectors are respectively the scanning wavelength, and the original mid-infrared spectrum absorption peak and the absorption peak of the second derivative spectrogram of the sample corresponding to the scanning wavelength. According to the classification result of the neural network model, the parvalbumin can be rapidly identified and detected. The invention does not need specific antibody or special reagent, test paper and instrument, can realize the purpose of rapid identification and detection of the allergen, and simultaneously realizes the requirement of no damage to the sample.
Description
Technical Field
The invention belongs to the technical field of biology, and relates to a method for quickly identifying and detecting major allergen parvalbumin in fish aquatic products.
Background
Parvalbumin is the most major allergen in freshwater fish, and causes increasingly complex and severe allergic symptoms. The traditional cooking method can not degrade parvalbumin, so food allergy events are often caused, and therefore, a method for rapidly identifying allergen parvalbumin is very necessary to be explored.
The current detection technology of allergen parvalbumin mainly comprises three types. The first is a molecular biology method, represented by Polymerase Chain Reaction (PCR) technology for detecting allergen-encoding DNA fragments; the second type is an immunological method, represented by enzyme linked immunosorbent assay (ELISA) for detecting allergenicity; the third type is an instrumental analysis method, which is represented by a mass spectrometry technology for analyzing the amino acid sequence of the whole protein or the peptide fragment after enzymolysis and a biosensor based on surface plasmon resonance.
Application No. 201510882721.4 discloses a method for screening a consensus gene sequence of fish parvalbumin by bioinformatics and comparative genomics, designing a specific amplification primer pair by using the sequence, and carrying out fluorescence quantitative PCR detection on a sample to be detected by using the primer pair; application No. 201110196761.5 discloses a monoclonal antibody prepared from a freshwater fish chub parvalbumin standard substance, which realizes trace detection of parvalbumin by competitive ELISA; application No. 201810432023.8 discloses a method for detecting aquatic allergen parvalbumin by liquid chromatography tandem mass spectrometry.
The mass spectrum technology has high cost and high operation requirement; the PCR detection technology is used for detecting the possible false negative result of the processed food; immunological detection techniques are based on antigen-antibody binding reactions, however, antibody preparation is difficult and time consuming.
In order to overcome the defects in the prior art, the invention aims to provide a method for quickly detecting parvalbumin by combining a mid-infrared spectrum technology and an inclusion-Resnet-v 2 neural network, which inputs an infrared spectrogram of a detected sample into a constructed model, can realize quick detection without an antibody and simultaneously realize the requirement of lossless food.
Disclosure of Invention
The invention aims to provide a method for rapidly identifying and detecting parvalbumin based on mid-infrared spectroscopy and neural network technology.
The technical scheme of the invention is that a method for rapidly identifying and detecting parvalbumin based on mid-infrared spectroscopy and neural network technology is characterized by comprising the following steps:
(1) performing infrared spectrum measurement on protein extracted from a sample to be measured to obtain original intermediate infrared spectrogram data;
(2) obtaining a second derivative map by performing second derivation on the obtained original intermediate infrared spectrogram;
(3) acquiring three-dimensional vector data from an original infrared spectrum and a second derivative spectrogram thereof, and identifying by using a trained neural network model to obtain a classification result: the three-dimensional vectors are respectively scanning wavelengths which correspond to the absorption of the original mid-infrared spectrum of the sample and the absorption of a second derivative spectrogram.
(4) According to the classification result of the neural network model, the parvalbumin can be rapidly identified and detected.
The sample to be detected is derived from fish aquatic products or fish aquatic products, or the sample to be detected contains fish aquatic products.
The preferable fish is any one or more of finless eel, crucian carp, grass carp, silver carp, black carp, mandarin fish, snakehead fish, tilapia, weever, eel, longsnout catfish, pelteobagrus fulvidraco, takifugu rubripes, takifugu flavidus and spotted silver carp.
In the step (1), the original intermediate infrared spectrum is obtained by preprocessing the original spectrum obtained by infrared spectrometry scanning through baseline correction, spectrum smoothing, spectrum calculation and normalization processing.
The scanning range of infrared spectrometry is 4000-650cm-1In the interval, the spectral resolution is 2-6 cm-1(ii) a The number of scanning times is 15-20, and the average spectrum is taken.
In a preferred embodiment of the invention, the conditions for the infrared spectrometric determination are: using single-point ATR method, testing temperature is 20-25 deg.C, humidity is within 45%, and spectral resolution is 4cm-1The measurement range is 4000-650cm-1The number of scans was 16 and the spectra were averaged.
Selecting 4000-650cm from an original middle infrared spectrogram and a second-order middle infrared spectrogram-1Is subjected to profile analysis.
In the step (3), preferably, the neural network model is inclusion-Resnet-v 2.
The extraction method of parvalbumin comprises the following steps:
I. stirring a sample to be detected, and then mixing the raw materials according to the weight ratio of 1 g: adding 4-10 mL of a pH 6.8-7.8 solution containing protease inhibitor and CaCl2Homogenizing the Tris-HCl buffer solution, centrifuging at low temperature, and taking supernatant a;
slowly adding trichloroacetic acid with the volume of 2-4% into the supernatant a, stirring in an ice bath, adjusting the pH to 5.2, centrifuging at low temperature, and taking the supernatant b;
and III, slowly adding trichloroacetic acid with the volume of 2-4% into the supernatant b, stirring in an ice bath, centrifuging at low temperature, and purifying the precipitate c.
In the step I, the protease inhibitor is PMSF (phenylmethylsulfonyl fluoride), and the content is 75-150 mmol/L, preferably 100 mmol/L. CaCl2The content of (b) is 5 to 20mmol/L, preferably 10 mmol/L.
And in the step III, purifying the precipitate c by electrophoresis detection and impurity removal, and freeze-drying.
The neural network model is trained by the following method:
(a) taking parvalbumin, aquatic product protein without parvalbumin and aquatic product protein containing parvalbumin (namely a mixture of the parvalbumin and the aquatic product protein), and carrying out infrared spectrum measurement on freeze-dried powder of different samples to obtain original middle infrared spectrum data;
(b) performing second-order derivation on the obtained original intermediate infrared spectrogram;
(c) acquiring three-dimensional vector data from an original infrared spectrum and a second derivative spectrogram of the original infrared spectrum, and inputting the acquired three-dimensional vector and a result of whether the acquired three-dimensional vector contains the parvalbumin into a neural network model for training; the three-dimensional vector is respectively the scanning wavelength, the infrared spectrum absorption corresponding to the scanning wavelength and the absorption of the second derivative spectrogram.
In the step (a), a parvalbumin pure product and a parvalbumin-free aquatic product protein are extracted from the freshwater fish aquatic product. The sources of each of the parvalbumin, the parvalbumin-free aquatic product protein or the parvalbumin-containing aquatic product protein are one or more kinds of freshwater fish aquatic products, preferably one to four kinds, and more preferably one or two kinds. The freshwater fish comprises finless eel, crucian carp, grass carp, silver carp, black carp, mandarin fish, snakehead fish, tilapia, weever, eel, longsnout catfish, pelteobagrus fulvidraco, takifugu rubripes, takifugu flavidus and spotted silver carp.
In the step (a), the original spectrum obtained by infrared spectrometry scanning is subjected to pretreatment operations of baseline correction, spectrum smoothing, spectrum calculation and normalization processing to obtain the original mid-infrared spectrum.
The scanning range of infrared spectrometry is 4000-650cm-1In the interval, the spectral resolution is2～6cm-1(ii) a The number of scanning times is 15-20, and the average spectrum is taken.
Preferably, the infrared spectroscopy conditions are: using single-point ATR method, testing temperature is 20-25 deg.C, humidity is within 45%, and spectral resolution is 4cm-1The measurement range is 4000-650cm-1The number of scans was 16 and the spectra were averaged.
Selecting 4000-650cm from an original middle infrared spectrogram and a second-order middle infrared spectrogram-1Is subjected to profile analysis.
In the step (c), preferably, the neural network model is inclusion-Resnet-v 2, and the learning method is transfer learning.
The number of neural network iterations is 50-200, and the software used is Python 3.7 and tensflow 1.10.0. The preferred number of iterations is 100.
The total recognition rate of the method for detecting the parvalbumin can reach more than 95 percent, the probability of false negative and false positive is low, and the omission factor is not more than 5 percent.
The invention overcomes the defects of the prior art, provides a method for rapidly detecting the main allergen parvalbumin in fish aquatic products based on a mid-infrared spectrum technology and a neural network, can realize rapid detection without an antibody, can achieve the purpose of rapid qualitative detection, and simultaneously realizes the requirement of no damage to samples.
The invention has the beneficial effects that:
(1) the infrared spectrum technology is used for detecting the parvalbumin, so that the testing process is simplified;
(2) after the neural network model is trained, the existence of the parvalbumin in the sample can be effectively identified, and the identification rate is high;
(3) the sample detection can be completed within a few seconds by adopting an analysis method of a mid-infrared spectrum and a neural network, and a result is obtained;
(4) the micro-sample can be tested, and the obtained protein sample is not damaged in the detection process, so that the extracted protein sample can be detected or utilized again;
(5) the method does not need specific antibodies or special reagents, test paper and instruments, and greatly reduces the cost, the operation difficulty and human errors compared with the existing method.
Drawings
FIG. 1 shows the IR absorption spectra (4000- & 650 cm) of parvalbumin from different sources and without parvalbumin in freshwater fish protein (sample of A, C groups) in example 1-1)。
FIG. 2 is an infrared second-order derivative spectrum (4000-650 cm) of parvalbumin from different sources and without parvalbumin in example 1 (samples from A, C panels)-1)。
Fig. 3 shows examples 2 and 3 comparing the results of the discrimination of different algorithms (neural network (IRN), Random Forest (RF), Support Vector Machine (SVM)). Pa represents the overall recognition accuracy, and groups A-C are the accuracy of the single Group of samples A-C respectively.
Detailed Description
Example 1
The method comprises the following steps of treating an aquatic product sample by using 16 kinds of fishes including finless eels, crucian carps, grass carps, silver carps, black carps, mandarin fish, snakeheads, tilapia, weever, eels, longsnout catfish, catfishes, pelteobagrus fulvidraco, takifugu rubripes and takifugu flavidus, firstly extracting parvalbumin and other proteins, wherein the specific steps of extracting the proteins are as follows:
(1) the fish white muscle was minced, and 5-fold volume of an extract (10mM CaCl2,100mM PMSF,10mM Tri-HCl, pH 7.5) was added and homogenized. Centrifuging at 4 deg.C and 10000rpm for 20min, and collecting precipitate A and supernatant a. Adding TCA (3%) slowly according to the volume of the supernatant, stirring in ice bath for 1h, adjusting pH to 5.2, centrifuging at 4 ℃ and 10000rpm for 20min, collecting precipitate B and supernatant B, repeating the above process, adding TCA, centrifuging at 4 ℃ and 10000rpm for 20min, and collecting precipitate C and supernatant C. Wherein the precipitate c is parvalbumin which is used as a relatively pure parvalbumin after electrophoresis detection and impurity removal.
(2) And (3) respectively collecting the twice precipitated A, B and the third supernatant C, determining the content of parvalbumin in the supernatant C by ELISA detection, and removing the parvalbumin for later use after multiple TCA precipitation, namely the freshwater fish protein without the parvalbumin.
And (2) mixing the 16 fish parvalbumin samples obtained in the step (1) in a pairwise equal volume manner, wherein the number of the samples can be increased to 136 (16 samples are single freshwater fish parvalbumin, and the other 120 samples are mixtures of any two freshwater fish parvalbumin) respectively), and the samples are set as a group A.
And (3) mixing the parts without parvalbumin of the 16 fishes obtained in the step (2) in pairs with equal volumes, wherein the number of samples can be enlarged to 136, and the samples are set as a group C, namely the samples without parvalbumin.
And mixing the corresponding samples of the group A and the group C in equal volume respectively to set the samples as a group B, namely the aquatic product protein sample containing the parvalbumin and other proteins.
Secondly, after all samples are frozen and dried, protein freeze-dried powder is taken to collect spectra. Spectral data acquisition was performed on each sample using a Spectrum GX fourier transform infrared spectrometer (Perkin Elmer corporation), and a solid measurement single point ATR accessory. The amount of lyophilized protein powder used was about 1 mg.
The spectrum collection mode and collection conditions are as follows: using single-point ATR method, testing temperature is 20-25 deg.C, humidity is within 45%, and spectral resolution is 4cm-1The measurement range is 4000-650cm-1The number of scans was 16.
And (3) performing pretreatment operations of baseline correction, Spectrum smoothing, Spectrum calculation and normalization on the acquired Spectrum by adopting Spectrum V3.02 operation software of a Perkin-Elmer company to obtain infrared spectrums of different samples. A. The original IR spectra of protein samples from various fish in group C are shown in FIG. 1. Most of the spectra in fig. 1 overlapped, indicating that there was no significant difference in the raw ir spectra for the various fish.
And performing second-order derivation on the infrared absorption spectrograms of the samples to obtain a second-order derivative spectrogram, wherein the second-order infrared spectrogram is shown in figure 2.
It can be seen from the primary and secondary ir spectra that the amide bands are not completely distinguishable depending on the distribution of the amide bands in the sample.
Example 2
The infrared spectrum and the second derivative spectrum of each sample obtained in example 1 are both a two-dimensional vector of 2 × 3351, and one of the dimensions of the two are the same, so that the data contained in each sample (the second order reflects the peaks masked in the original) can be represented by a three-dimensional vector of 3 × 3351.
A. And (3) extracting 70% (96) of samples in each group of the samples B and C to serve as a training set, inputting corresponding 3X 3550 vectors and corresponding results (group) for training the constructed increment-Resnet-v 2 neural network model, and using the remaining 30% (40) of samples as a verification set for verifying the recognition accuracy of the model, wherein the results are shown in FIG. 3(IRN) and Table 1.
TABLE 1 IRN identification Rate
The classification results for this sample set against a Support Vector Machine (SVM) model are shown in fig. 3 (SVM). The neural network only has a false negative, the recognition rate of the neural network on a sample containing the parvalbumin reaches 98.8 percent, the recognition rate of a sample without the parvalbumin is 95 percent, and the accuracy (Pa) is 97.5 percent; 87.5%, 75% and 83.3% higher than those of the support vector machine. Meanwhile, for the recognition rate of each group, the neural network is higher than the neural network, so that the rapid recognition and identification of the parvalbumin are realized.
Example 3
Still using the neural network classification results in example 2, the classification results of the comparative Random Forest (RF) algorithm on the samples obtained in example 1 are shown in fig. 3(RF), and the recognition rate of the random forest on the sample containing parvalbumin is 85%, the recognition rate of the sample without parvalbumin is 85%, the accuracy (Pa) is 93.3%, and is also lower than the recognition result of the neural network. Meanwhile, the recognition rate for each group is relatively low, especially for the recognition of the group A.
Examples 2 and 3 both show that the use of mid-infrared spectroscopy and neural network technology allows for the rapid identification and detection of fish allergen parvalbumin.
Claims (9)
1. A method for rapidly identifying and detecting parvalbumin based on mid-infrared spectroscopy and neural network technology is characterized by comprising the following steps:
(1) extracting protein from a sample to be detected, and performing infrared spectrum measurement to obtain original intermediate infrared spectrum data;
(2) obtaining a second derivative map by performing second derivation on the obtained original intermediate infrared spectrogram;
(3) acquiring three-dimensional vector data from an original infrared spectrum and a second derivative spectrogram thereof, wherein the three-dimensional vectors are scanning wavelengths, and an absorption peak of the original mid-infrared spectrum of a sample corresponding to the scanning wavelengths and an absorption peak of the second derivative spectrogram; identifying by using the trained neural network model to obtain a classification result;
(4) rapidly identifying and detecting parvalbumin according to the classification result of the neural network model;
the neural network model is trained by the following method:
(a) taking parvalbumin, aquatic product protein without parvalbumin and aquatic product protein containing parvalbumin, and carrying out infrared spectrum measurement on the different samples to obtain original intermediate infrared spectrum data;
(b) performing second-order derivation on the obtained original intermediate infrared spectrogram;
(c) acquiring three-dimensional vector data from an original infrared spectrum and a second derivative spectrogram thereof, wherein the three-dimensional vectors are respectively a scanning wavelength, an absorption peak of the original mid-infrared spectrum of a sample corresponding to the scanning wavelength and an absorption peak of the second derivative spectrogram; and (4) inputting the obtained three-dimensional vector and the result of judging whether the three-dimensional vector contains the parvalbumin as a training set into a neural network model for training.
2. The method for rapidly identifying and detecting parvalbumin based on mid-infrared spectroscopy and neural network technology according to claim 1, wherein the sample to be detected is derived from fish aquatic products or fish aquatic products, or the sample to be detected contains fish aquatic products.
3. The method for rapidly identifying and detecting parvalbumin based on mid-infrared spectroscopy and neural network technology according to claim 2, wherein the fish aquatic product is any one or more of finless eel, crucian carp, grass carp, silver carp, black carp, mandarin fish, snakehead, tilapia, weever, eel, longsnout catfish, pelteobagrus fulvidraco, fugu rubripes, fugu flavidus and spotted silver carp.
4. The method for rapidly identifying and detecting parvalbumin based on mid-infrared spectroscopy and neural network technology as claimed in claim 1, wherein in step (1), the original spectrum obtained by the infrared spectrometric scanning is subjected to preprocessing operations of baseline correction, spectrum smoothing, spectrum calculation and normalization to obtain the original mid-infrared spectrum.
5. The method for rapid identification and detection of parvalbumin based on mid-infrared spectroscopy and neural network technology according to claim 1, wherein the conditions of infrared spectrometry in step (1) and step (a) are as follows: the scanning range is 4000-650cm-1In the interval, the spectral resolution is 2-6 cm-1(ii) a The number of scanning times is 15-20, and the average spectrum is taken.
6. The method for rapid identification and detection of parvalbumin based on mid-infrared spectroscopy and neural network technology as claimed in claim 1, wherein in step (1) and step (a), the infrared spectroscopy is performed under the following conditions: using single-point ATR method, testing temperature is 20-25 deg.C, humidity is within 45%, and spectral resolution is 4cm-1The number of scans was 16, and the spectra were averaged.
7. The method for rapidly identifying and detecting parvalbumin based on mid-infrared spectroscopy and neural network technology as claimed in claim 1, wherein in step (3) and step (c), 4000-650cm is selected from original mid-infrared spectrogram and second-order mid-infrared spectrogram-1Is subjected to profile analysis.
8. The method for rapid identification and detection of parvalbumin based on mid-infrared spectroscopy and neural network technology according to claim 1, wherein the neural network model is inclusion-Resnet-v 2.
9. The method for rapid identification and detection of parvalbumin based on mid-infrared spectroscopy and neural network technology as claimed in claim 1, wherein the training of neural network model is based on transfer learning.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201811455680.0A CN109253983B (en) | 2018-11-30 | 2018-11-30 | Method for rapidly identifying and detecting parvalbumin based on mid-infrared spectrum and neural network technology |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201811455680.0A CN109253983B (en) | 2018-11-30 | 2018-11-30 | Method for rapidly identifying and detecting parvalbumin based on mid-infrared spectrum and neural network technology |
Publications (2)
| Publication Number | Publication Date |
|---|---|
| CN109253983A CN109253983A (en) | 2019-01-22 |
| CN109253983B true CN109253983B (en) | 2021-07-27 |
Family
ID=65042310
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN201811455680.0A Active CN109253983B (en) | 2018-11-30 | 2018-11-30 | Method for rapidly identifying and detecting parvalbumin based on mid-infrared spectrum and neural network technology |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN109253983B (en) |
Families Citing this family (5)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN110057771A (en) * | 2019-06-13 | 2019-07-26 | 四川德成动物保健品有限公司 | A kind of discrimination method of viola mandshurica and dandelion |
| CN112362611A (en) * | 2020-10-30 | 2021-02-12 | 上海海洋大学 | Method for rapidly and qualitatively detecting chemical components in marinated broth |
| CN113699256B (en) * | 2021-09-26 | 2023-06-20 | 浙江省农业科学院 | Functional marker for rapidly identifying gynogenesis offspring of salmon and koi and application of functional marker |
| CN113945537A (en) * | 2021-09-27 | 2022-01-18 | 桂林电子科技大学 | High-accuracy near infrared spectrum quantitative model establishing method |
| CN114062305B (en) * | 2021-10-15 | 2024-01-26 | 中国科学院合肥物质科学研究院 | Single grain variety identification method and system based on near infrared spectrum and 1D-In-Resnet network |
Citations (2)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN102707064A (en) * | 2011-07-14 | 2012-10-03 | 集美大学 | Method for sensitive detection of freshwater fish main allergen parvalbumin |
| CN105911127A (en) * | 2016-04-14 | 2016-08-31 | 浙江工商大学 | Rapid detection method of fish parvalbumin by capillary electrophoresis |
Family Cites Families (6)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| JPH0829335A (en) * | 1994-07-15 | 1996-02-02 | Kubota Corp | Rice analyzing and evaluating apparatus |
| US7283228B2 (en) * | 2003-04-11 | 2007-10-16 | Purdue Research Foundation | Process and apparatus for segregation and testing by spectral analysis of solid deposits derived from liquid mixtures |
| CN101251526B (en) * | 2008-02-26 | 2012-08-29 | 浙江大学 | Method and apparatus for nondestructively testing food synthetic quality |
| CN101339186A (en) * | 2008-08-07 | 2009-01-07 | 中国科学院过程工程研究所 | Method for on-line detection for solid-state biomass bioconversion procedure |
| CN103245629A (en) * | 2013-04-27 | 2013-08-14 | 丽水学院 | Method for identifying Dendrobium aphyllum (Roxb.)C. E. Fisch., Dendrobium officinale Kimura et Migo and Dendrobium devonianum Paxt. |
| CN107478595B (en) * | 2017-08-14 | 2020-09-15 | 上海海洋大学 | Method for rapidly identifying authenticity of pearl powder and quantitatively predicting content of adulterated shell powder |
-
2018
- 2018-11-30 CN CN201811455680.0A patent/CN109253983B/en active Active
Patent Citations (2)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN102707064A (en) * | 2011-07-14 | 2012-10-03 | 集美大学 | Method for sensitive detection of freshwater fish main allergen parvalbumin |
| CN105911127A (en) * | 2016-04-14 | 2016-08-31 | 浙江工商大学 | Rapid detection method of fish parvalbumin by capillary electrophoresis |
Non-Patent Citations (2)
| Title |
|---|
| 基于中红外光谱技术的鱼类过敏原小清蛋白的快速鉴别与定量检测;张晓鹏;《中国优秀硕士学位论文全文数据库 工程科技Ⅰ辑》;20200331(第3期);第B014-795页 * |
| 食物过敏原构象性表位鉴别的研究进展;李欣;《食品科学》;20120915;第33卷(第17期);第279-283页 * |
Also Published As
| Publication number | Publication date |
|---|---|
| CN109253983A (en) | 2019-01-22 |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| CN109253983B (en) | Method for rapidly identifying and detecting parvalbumin based on mid-infrared spectrum and neural network technology | |
| Ortea et al. | Review on proteomics for food authentication | |
| Zia et al. | Current analytical methods for porcine identification in meat and meat products | |
| Jaiswal et al. | Detection and quantification of soymilk in cow–buffalo milk using Attenuated Total Reflectance Fourier Transform Infrared spectroscopy (ATR–FTIR) | |
| Yasui et al. | An automated peak identification/calibration procedure for high‐dimensional protein measures from mass spectrometers | |
| Lin et al. | A bioinformatics approach to identify patients with symptomatic peanut allergy using peptide microarray immunoassay | |
| Valletta et al. | Mass spectrometry-based protein and peptide profiling for food frauds, traceability and authenticity assessment | |
| Liu et al. | Rapid discrimination of three marine fish surimi by Tri-step infrared spectroscopy combined with Principle Component Regression | |
| CN106124445A (en) | A kind of quick, Undamaged determination genetically engineered soybean method | |
| CN111766323B (en) | Characteristic peptide combination and method for detecting milk doped in camel milk | |
| CN101196526A (en) | Mass spectrometry reagent kit and method for rapid tuberculosis diagnosis | |
| Rochfort et al. | Mussel metabolomics—Species discrimination and provenance determination | |
| CN111678973A (en) | Method for rapidly identifying Atlantic salmon and rainbow trout based on mass spectrometry technology | |
| CN111766324A (en) | Characteristic peptide combination and method for detecting milk doped in buffalo milk | |
| Zhao et al. | Identification of donkey meat in foods using species-specific PCR combined with lateral flow immunoassay | |
| CN118112160A (en) | Method for identifying 11 allergic proteins in silkworm chrysalis | |
| Azkargorta et al. | TUBEs-mass spectrometry for identification and analysis of the ubiquitin-proteome | |
| CN113310934A (en) | Method for quickly identifying milk cow milk mixed in camel milk and mixing proportion thereof | |
| CN109557193B (en) | Mass spectrum qualitative detection method for main sesame allergen | |
| Magdas et al. | Botanical honey recognition and quantitative mixture detection based on Raman spectroscopy and machine learning | |
| Shi et al. | Rapid authentication of Indonesian edible bird's nests by near-infrared spectroscopy and chemometrics | |
| CN115561192A (en) | Multistage infrared spectrum rapid evaluation method for quality of protein peptide powder | |
| CN109001144A (en) | Method based on middle infrared spectrum Rapid identification and detection tropomyosin | |
| CN108676863B (en) | Method for identifying celiac allergen in wheat flour by high-throughput sequencing technology | |
| CN105738313A (en) | Method for identifying animal blood on basis of near-infrared spectrum technologies and application of method |
Legal Events
| Date | Code | Title | Description |
|---|---|---|---|
| PB01 | Publication | ||
| PB01 | Publication | ||
| SE01 | Entry into force of request for substantive examination | ||
| SE01 | Entry into force of request for substantive examination | ||
| GR01 | Patent grant | ||
| GR01 | Patent grant |