CN116183924B - Serum metabolism marker for liver cancer risk prediction and screening method and application thereof - Google Patents
Serum metabolism marker for liver cancer risk prediction and screening method and application thereof Download PDFInfo
- Publication number
- CN116183924B CN116183924B CN202310451383.3A CN202310451383A CN116183924B CN 116183924 B CN116183924 B CN 116183924B CN 202310451383 A CN202310451383 A CN 202310451383A CN 116183924 B CN116183924 B CN 116183924B
- Authority
- CN
- China
- Prior art keywords
- liver cancer
- acid
- risk prediction
- cancer risk
- serum
- 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
Images
Classifications
-
- G—PHYSICS
- G01—MEASURING; TESTING
- G01N—INVESTIGATING OR ANALYSING MATERIALS BY DETERMINING THEIR CHEMICAL OR PHYSICAL PROPERTIES
- G01N33/00—Investigating or analysing materials by specific methods not covered by groups G01N1/00 - G01N31/00
- G01N33/48—Biological material, e.g. blood, urine; Haemocytometers
- G01N33/50—Chemical analysis of biological material, e.g. blood, urine; Testing involving biospecific ligand binding methods; Immunological testing
- G01N33/53—Immunoassay; Biospecific binding assay; Materials therefor
- G01N33/574—Immunoassay; Biospecific binding assay; Materials therefor for cancer
- G01N33/57407—Specifically defined cancers
- G01N33/57438—Specifically defined cancers of liver, pancreas or kidney
-
- G—PHYSICS
- G01—MEASURING; TESTING
- G01N—INVESTIGATING OR ANALYSING MATERIALS BY DETERMINING THEIR CHEMICAL OR PHYSICAL PROPERTIES
- G01N30/00—Investigating or analysing materials by separation into components using adsorption, absorption or similar phenomena or using ion-exchange, e.g. chromatography or field flow fractionation
- G01N30/02—Column chromatography
-
- G—PHYSICS
- G01—MEASURING; TESTING
- G01N—INVESTIGATING OR ANALYSING MATERIALS BY DETERMINING THEIR CHEMICAL OR PHYSICAL PROPERTIES
- G01N30/00—Investigating or analysing materials by separation into components using adsorption, absorption or similar phenomena or using ion-exchange, e.g. chromatography or field flow fractionation
- G01N30/02—Column chromatography
- G01N30/62—Detectors specially adapted therefor
- G01N30/72—Mass spectrometers
-
- G—PHYSICS
- G16—INFORMATION AND COMMUNICATION TECHNOLOGY [ICT] SPECIALLY ADAPTED FOR SPECIFIC APPLICATION FIELDS
- G16B—BIOINFORMATICS, i.e. INFORMATION AND COMMUNICATION TECHNOLOGY [ICT] SPECIALLY ADAPTED FOR GENETIC OR PROTEIN-RELATED DATA PROCESSING IN COMPUTATIONAL MOLECULAR BIOLOGY
- G16B40/00—ICT specially adapted for biostatistics; ICT specially adapted for bioinformatics-related machine learning or data mining, e.g. knowledge discovery or pattern finding
-
- G—PHYSICS
- G16—INFORMATION AND COMMUNICATION TECHNOLOGY [ICT] SPECIALLY ADAPTED FOR SPECIFIC APPLICATION FIELDS
- G16B—BIOINFORMATICS, i.e. INFORMATION AND COMMUNICATION TECHNOLOGY [ICT] SPECIALLY ADAPTED FOR GENETIC OR PROTEIN-RELATED DATA PROCESSING IN COMPUTATIONAL MOLECULAR BIOLOGY
- G16B50/00—ICT programming tools or database systems specially adapted for bioinformatics
- G16B50/30—Data warehousing; Computing architectures
-
- G—PHYSICS
- G16—INFORMATION AND COMMUNICATION TECHNOLOGY [ICT] SPECIALLY ADAPTED FOR SPECIFIC APPLICATION FIELDS
- G16H—HEALTHCARE INFORMATICS, i.e. INFORMATION AND COMMUNICATION TECHNOLOGY [ICT] SPECIALLY ADAPTED FOR THE HANDLING OR PROCESSING OF MEDICAL OR HEALTHCARE DATA
- G16H50/00—ICT specially adapted for medical diagnosis, medical simulation or medical data mining; ICT specially adapted for detecting, monitoring or modelling epidemics or pandemics
- G16H50/30—ICT specially adapted for medical diagnosis, medical simulation or medical data mining; ICT specially adapted for detecting, monitoring or modelling epidemics or pandemics for calculating health indices; for individual health risk assessment
-
- Y—GENERAL TAGGING OF NEW TECHNOLOGICAL DEVELOPMENTS; GENERAL TAGGING OF CROSS-SECTIONAL TECHNOLOGIES SPANNING OVER SEVERAL SECTIONS OF THE IPC; TECHNICAL SUBJECTS COVERED BY FORMER USPC CROSS-REFERENCE ART COLLECTIONS [XRACs] AND DIGESTS
- Y02—TECHNOLOGIES OR APPLICATIONS FOR MITIGATION OR ADAPTATION AGAINST CLIMATE CHANGE
- Y02A—TECHNOLOGIES FOR ADAPTATION TO CLIMATE CHANGE
- Y02A90/00—Technologies having an indirect contribution to adaptation to climate change
- Y02A90/10—Information and communication technologies [ICT] supporting adaptation to climate change, e.g. for weather forecasting or climate simulation
Abstract
The invention belongs to the technical field of biomedicine, and relates to a serum metabolic marker for liver cancer risk prediction, a screening method and application thereof. The invention provides a serum metabolic marker for liver cancer risk prediction, which comprises at least one of the following compounds: 3-hydroxybutyric acid; cystine; lactate; erucic acid; oleoylethanolamide; ethyl malonate; 3-hydroxy-2-propionic acid ethyl ester; hexadecanedioic acid; niu Huangdan cholic acid-3-sulfate; tryptophan betaine; tridecendioic acid; octadecenodiacyl carnitines; taurodeoxycholic acid-3-sulfate; ceramide; n, N-trimethyl-5-aminopentanoic acid; 3-hydroxydecanoyl carnitine; 3-hydroxyoctanoyl carnitine. The invention shows that the specific serum metabolite can be used as a novel minimally invasive biomarker, the disease risk prediction level is improved, and the successful development of the biomarker provides a method and a strategic reference for the development of other disease biomarkers.
Description
Technical Field
The invention belongs to the technical field of biomedicine, and particularly relates to a serum metabolic marker for liver cancer risk prediction, a screening method and application thereof.
Background
Primary hepatocellular carcinoma (HCC) is one of the most common primary malignancies worldwide. The disease stage in the diagnosis of HCC largely determines the efficacy of subsequent treatments. At present, the traditional liver cancer markers (such as alpha fetoprotein AFP) applied clinically also lack satisfactory sensitivity and specificity in diagnosis, so that more reliable and accurate tumor markers are urgently needed to be searched for in order to predict the risk of liver cancer transfer of high-risk groups and improve the chance of early-stage liver cancer discovery.
Metabonomics, which is a branch of multiple groups, is a systematic study that uses high-throughput techniques to perform qualitative and quantitative analysis of small molecule metabolites, such as amino acids, nucleotides, carbohydrates, lipids, etc., with a relative molecular mass of <1000×103 in the living system. Metabonomics typically uses techniques based on nuclear magnetic resonance, liquid chromatography-mass spectrometry, and gas chromatography-mass spectrometry, etc., which not only recognize complex metabolic phenotypes, but also integrate bioinformatic data and other histologic strategies such as genomics, transcriptomics, and proteomics to elucidate the underlying biological mechanisms of the disease and to discover clinically relevant diagnostic and prognostic markers of disease risk. The time response of metabolomics is the fastest among all groups, and the results are easily understood, easily combined with phenotypes and functions, and have been widely used in the fields of disease diagnosis, therapy monitoring and pharmacodynamic evaluation.
In recent years, metabonomics is gradually rising, new progress is made in research in multiple fields such as occurrence mechanisms of diseases such as cancers, early prevention and diagnosis, drug treatment targets and the like, and a plurality of serum metabolite indexes with excellent sensitivity and specificity are developed, so that the defects of the traditional serological indexes are overcome, and early screening and early diagnosis of the diseases such as cancers are assisted.
Disclosure of Invention
The invention aims to solve the technical problem of providing a serum metabolic marker for liver cancer risk prediction, and a screening method and application thereof, so as to find a more reliable and more accurate tumor marker to predict the risk of liver cancer transfer of high-risk groups and improve the chance of early liver cancer discovery.
In order to solve the technical problems, the invention adopts the following technical scheme:
a serum metabolic marker for liver cancer risk prediction, the serum metabolic marker comprising at least one of the following compounds:
3-hydroxybutyric acid;
cystine;
lactate;
erucic acid;
oleoylethanolamide;
ethyl malonate;
3-hydroxy-2-propionic acid ethyl ester;
hexadecanedioic acid;
niu Huangdan cholic acid-3-sulfate;
tryptophan betaine;
tridecendioic acid;
octadecenodiacyl carnitines;
taurodeoxycholic acid-3-sulfate;
ceramide;
n, N-trimethyl-5-aminopentanoic acid;
3-hydroxydecanoyl carnitine;
3-hydroxyoctanoyl carnitine.
Optionally, the serum metabolic marker comprises at least one of the following compounds:
cystine;
lactate;
oleoylethanolamide;
niu Huangdan cholic acid-3-sulfate;
octadecenodiacyl carnitines;
taurodeoxycholic acid-3-sulfate;
ceramide;
n, N-trimethyl-5-aminopentanoic acid;
3-hydroxy-2-propionic acid ethyl ester;
hexadecanedioic acid;
3-hydroxydecanoyl carnitine.
Optionally, the serum metabolic marker comprises the following compounds:
cystine;
lactate;
oleoylethanolamide;
niu Huangdan cholic acid-3-sulfate;
octadecenodiacyl carnitines;
taurodeoxycholic acid-3-sulfate;
ceramide;
n, N, N-trimethyl-5-aminopentanoic acid.
Optionally, the serum metabolic marker comprises the following compounds:
cystine;
oleoylethanolamide;
3-hydroxy-2-propionic acid ethyl ester;
hexadecanedioic acid;
taurodeoxycholic acid-3-sulfate;
ceramide;
3-hydroxydecanoyl carnitine.
The invention also provides application of the serum metabolic marker product for liver cancer risk prediction in preparing the liver cancer risk prediction product.
Optionally, the product for detecting the serum metabolic marker for liver cancer risk prediction and/or the product for liver cancer risk prediction comprises a reagent, test paper, a kit or an instrument.
The invention also provides a screening method of serum metabolic markers for liver cancer risk prediction, which comprises the following steps:
step one: collecting a blood sample of a subject at risk of developing liver cancer, and dividing the blood sample into a cancer-transferring group and a non-cancer-transferring group according to whether the subject progresses to liver cancer;
step two: determining the metabolite content in each blood sample, and analyzing the relativity of the determined metabolite content with whether the liver cancer is progressed or not by using 3 or more than 3 statistical methods; setting a correlation threshold of each statistical method;
and thirdly, screening out metabolites at least reaching a correlation threshold of one statistical method.
Optionally, the statistical method comprises: variable importance projection and difference multiple; the Boruta method; lasso method; extraTree method; or the SelectKBest method;
the correlation threshold value of each statistical method is respectively that the variable importance projection is >1 and the difference multiple is >1.5; ten random tests are carried out by adopting a Borata method, and the first 20 is sorted according to the feature selection times; performing ten random experiments by using Lasso, and sequencing the first 20 according to the feature selection times; performing ten random tests by adopting ExtraTree, and sequencing the first 20 according to the feature selection times; ten random trials were performed using a SelectKBest, ranked 20 before feature selection.
Alternatively, the blood sample of the subject at risk of developing liver cancer is a blood sample of a patient suffering from hepatitis b or cirrhosis and not yet progressing to liver cancer.
Optionally, in the second step, the metabolite content in each blood sample is determined by means of liquid chromatography-mass spectrometry.
The invention also provides a liver cancer risk prediction method, which comprises the following steps:
step a: establishing a liver cancer risk prediction model by utilizing the serum metabolic markers for liver cancer risk prediction;
step b: determining the serum metabolic marker content in a blood sample of the subject;
step c: and calculating a liver cancer risk prediction result through the liver cancer risk prediction model according to the content of the determined serum metabolic marker.
Alternatively, the serum metabolic marker combination for liver cancer risk prediction is cystine, lactate, oleoylethanolamine, taurocholate-3-sulfate, octadecenodiacyl carnitine, taurodeoxycholate-3-sulfate, ceramide carnitine, N, N, N-trimethyl-5-aminopentanoic acid or cystine, oleoylethanolamine, 3-hydroxy-2-ethyl propionate, hexadecanedioic acid, taurodeoxycholate-3-sulfate, ceramide, 3-hydroxydecanoyl carnitine.
Alternatively, the serum metabolic marker combinations for liver cancer risk prediction are: cystine; lactate; oleoylethanolamide; niu Huangdan cholic acid-3-sulfate; octadecenodiacyl carnitines; taurodeoxycholic acid-3-sulfate; ceramide; n, N, N-trimethyl-5-aminopentanoic acid. The logistic model established using the combination of the 8 serum metabolic markers described above was y= -0.983 log (cystine abundance) -2.951 log (lactate abundance) +0.878 log (oleoylethanolamine abundance) -0.568 log (Niu Huangdan cholic acid-3-sulfate abundance) +1.163 log (octadecendiolcholine abundance) +1.361 log (taurodeoxycholic acid-3-sulfate abundance) +1.189 log (ceramide abundance) +0.967 log (N, N-trimethyl-5-aminopentanoic acid) +1.256. When the value of y > is-0.43, it is determined that the subject is at high risk of progressing to liver cancer.
Alternatively, the serum metabolic marker combinations for liver cancer risk prediction are: cystine; oleoylethanolamide; 3-hydroxy-2-propionic acid ethyl ester; hexadecanedioic acid; taurodeoxycholic acid-3-sulfate; ceramide; 3-hydroxydecanoyl carnitine. The gradient lifting tree model established by the combination of the 7 serum metabolic markers is used, the iteration times are 75, the number of randomly extracted features is 100% when the tree is established, the minimum weight used when the tree is lifted is 1, the learning rate is 0.01, the minimum loss required when a leaf partition is newly added is 0.5, the sub-sample data occupy 50% of the whole observed proportion, and the maximum depth of the tree is 2. Substituting the detected abundance of each substance into a gradient lifting tree model calculated value, and judging that the subject has high risk of developing liver cancer when the value is more than 0.51.
Compared with the prior art, the invention has the beneficial effects that:
(1) According to the invention, 17 metabolites are screened out to be related to the incidence risk of liver cancer, a group of 8 serum metabolites and a group of 7 serum metabolites are determined and used for predicting the incidence risk of liver cancer, and good sensitivity and specificity are shown, so that a new technical support is provided for identifying high-risk groups of liver cancer and early diagnosis and treatment;
(2) The invention provides a screening method of serum metabolic markers for liver cancer risk prediction, which develops nest type case comparison research of liver cancer, and utilizes a plurality of statistical methods to analyze and determine the relativity between the content of the metabolite and whether the metabolite is progressed into liver cancer
(3) The invention shows that the specific serum metabolite can be used as a novel micro-invasive biomarker to improve the disease risk prediction level, and the successful development of the biomarker provides a method and a strategic reference for the development of other disease biomarkers.
Drawings
FIG. 1 is a statistical plot of the content of 17 metabolites in serum of peripheral blood; wherein, p < 0.05, p < 0.01, p < 0.001 are statistically different in the group with metastatic cancer compared to the group with non-metastatic cancer.
FIG. 2 is a graph of subject performance profile analysis results for 8 and 7 metabolites; wherein A is the result of the logistic regression model in the training set and 100 resampling (boottrap) internal verifications; b is the result of the gradient lifting tree model in the training set and the verification set.
Detailed Description
For a better description of the objects, technical solutions and advantages of the present invention, the present invention will be further described with reference to the following specific examples. It will be appreciated by persons skilled in the art that the specific embodiments described herein are for purposes of illustration only and are not intended to be limiting.
The test methods used in the examples are conventional methods unless otherwise specified; the materials, reagents and the like used, unless otherwise specified, are all commercially available.
Example 1
The embodiment provides a screening method of serum metabolic markers for liver cancer risk prediction, which comprises the following steps:
step one: blood samples of subjects at risk of developing liver cancer are collected, and the blood samples are classified into a cancer-transferring group and a non-cancer-transferring group according to whether or not the subjects progress to liver cancer. In the step, a unified standard queue specimen library and a database are established: standard-compliant blood samples were collected with standard procedures (SOP) and the system collected complete demographic and clinical data. A nationwide multicenter CHB patient was included in a long-term follow-up study of cases where hepatitis b or cirrhosis progressed to liver cancer, hepatitis b or cirrhosis did not progress to liver cancer. The serum specimens of the liver cancer patients in the last 2 years before the diagnosis and the serum specimens of the non-liver cancer patients in the last 2 years before the follow-up are matched with the clinical characteristics (common liver cancer related risk factors such as age, sex, alpha fetoprotein and the like) of the two groups of patients in the last 2 years. For a blood sample of a liver cancer patient before 2 years of definitive diagnosis and a non-liver cancer patient before 2 years of last follow-up, a non-target metabonomics technology is utilized to screen and verify metabolic markers related to liver cancer risks.
Step two: the metabolite content in each blood sample was measured by using an ultra-high performance liquid chromatograph-mass spectrometer, and 1033 named metabolites were measured in the serum in this example. Using 5 statistical methods to analyze the relativity of the measured metabolite content and whether the liver cancer happens; defining a correlation threshold of each statistical method;
and thirdly, screening out metabolites at least reaching a correlation threshold of one statistical method.
The implementation operation process comprises the following steps:
1. selection of study samples:
56 new liver cancer patients from the prospective study cohort were collected, matching 30 liver cirrhosis patients and 31 hepatitis b patients.
Setting test groups: non-metastatic cancer group: namely NHCC group comprising peripheral blood of patients with liver cirrhosis, peripheral blood of patients with hepatitis B2 years before last follow-up, 61 cases; cancer group: i.e., HCC group, liver cancer patients were diagnosed with peripheral blood, 56 cases 2 years ago.
The study was performed with 117 standard-compliant samples.
2. Extraction of serum samples:
each subject used procoagulant blood collection tubes to collect morning-drawn fasting venous blood at a rate of 5ml x 2 tubes/person. All subjects sign sample collection informed consent, and the processed samples follow the laboratory biosafety operation standard and make corresponding records.
Samples were prepared using an automatic MicroLab STAR system from Hamilton corporation. Several recovery standards were added for quality control prior to the first step of the extraction process. To remove proteins, dissociate small molecules bound to proteins or captured by the precipitated protein matrix, and recover chemically diverse metabolites, proteins were precipitated with methanol under vigorous shaking for 2 min (Glen Mills GenoGrinder) and then centrifuged. The obtained extract is divided into five parts: two of which are used for analysis by two different Reverse Phase (RP)/UPLC-MS/MS methods, using positive ion mode electrospray ionization (ESI), one for using negative ion mode ESI, one for analysis using HILIC/UPLC-MS/MS, and one for reserving one sample. The samples were briefly placed on a TurboVap cube (Zymark) to remove the organic solvent. The sample extracts were stored overnight under nitrogen before being ready for analysis.
3. Metabonomics detection:
waters ACQUITY ultra-high Performance liquid chromatography (UPLC) and Thermo Scientific Q-exact high resolution/precision mass spectrometers were used. The mass spectrometer was connected to a heated electrospray ionization (HESI-II) source and an Orbitrap mass analyzer operating at a mass resolution of 35,000. The sample extract is dried and then reconstituted in a solvent compatible with each of the four methods. Each reconstituted solvent contained a series of standards at fixed concentrations to ensure injection and chromatographic consistency. An aliquot was analyzed using acidic positive ion conditions and the chromatography was optimized to obtain more hydrophilic compounds. In this method, the extract is eluted from a C18 column (Waters UPLC BEH C18-2.1X100 mm,1.7 μm) gradient using water and methanol containing 0.05% perfluorooctanoic acid (PFPA) and 0.1% Formic Acid (FA). Another aliquot was also analyzed using acidic positive ion conditions, but was chromatographically optimized to obtain a more hydrophobic compound. In this method, the extract was eluted from the same C18 column as described above using a gradient of methanol, acetonitrile, water, 0.05% pfpa and 0.01% fa, and operated at an overall higher organic content. Another aliquot was analyzed using a separate dedicated C18 column using basic anion optimization conditions. The alkaline extract was eluted from the column using a gradient of methanol and water, but using 6.5mM ammonium bicarbonate at pH 8. The fourth aliquot was analyzed by negative ion after elution from the HILIC column (Waters UPLC BEH Amide 2.1.1X1150 mM,1.7 μm) using a gradient consisting of water and acetonitrile with 10mM ammonium formate (pH 10.8). MS analysis using dynamic exclusion in MS and data-dependent MS n Alternating between scans. The scan range does not vary much between different methods, but covers 70-1000 m/z.
4. And (3) data processing:
performing peak identification, peak extraction, peak alignment, integration and other treatments, and performing substance annotation to determine 1033 named metabolites in serum at a plurality of time points; filtering the outliers based on the relative standard deviation (relative standard deviation); filling the missing data by adopting a minimum value; normalization was performed using an internal standard (internal standard).
5. Statistical analysis:
based on the abundance of each metabolite tested, the differential metabolites were screened by the following 5 methods: (1) Variable Importance Projection (VIP) and fold difference (FC): performing partial least squares discriminant analysis and difference multiple analysis by adopting a MetaboAnalyst online website, and screening and distinguishing differential metabolites of liver cancer and non-liver cancer by taking VIP >1 and FC >1.5 as standards; (2) Boruta: the feature importance is calculated by extracting features and disturbing the sequence of the features by adopting a random forest method based on a feature selection method carried out by a Borata package in R; (3) Lasso: a feature selection method based on the glrnet packet in R; (4) ExtraTree: an extremely random tree based on python operation is composed of a plurality of decision trees and is characterized by random characteristics, random parameters, random models and random splitting; (5) SelectKBest: using python operation, the method uses a feature selection function by scoring features and then selecting features from high to low.
The metabolite selected must meet 3 or more conditions or play an important role in the subsequent modeling process: (1) VIP >1 and FC >1.5; (2) Performing ten random tests by using Borata, and sequencing the first 20 according to the feature selection times; (3) Performing ten random experiments by using Lasso, and sequencing the first 20 according to the feature selection times; (4) Performing ten random tests by adopting ExtraTree, and sequencing the first 20 according to the feature selection times; (5) Ten random tests are carried out by adopting SelectKBest, and the first 20 is sorted according to the feature selection times; based on the above conditions, the 17 metabolites were screened for utility in differentiating between the metastatic and non-metastatic groups.
17 significantly different metabolites were obtained compared to the non-metastatic case, see table 1 and figure 1. The Mann-Whitney U test is used in fig. 1, wherein x, x represent that the metastatic cancer group has a statistical difference p < 0.05, p < 0.01, p < 0.001 compared to the non-metastatic cancer group.
TABLE 1
In summary, the invention utilizes prospective crowd queues to develop nest type case control research, detects 1033 named metabolites in serum through a non-targeted metabonomics technology, adopts 5 methods to perform feature selection (MetaboAnalyst, boruta, lasso, extraTree, selectKBest), and finds that 17 metabolites are obviously related to the pathogenesis of liver cancer, and specifically comprises the following steps:
3-hydroxybutyric acid;
cystine;
lactate;
erucic acid;
oleoylethanolamide;
ethyl malonate;
3-hydroxy-2-propionic acid ethyl ester;
hexadecanedioic acid;
niu Huangdan cholic acid-3-sulfate;
tryptophan betaine;
tridecendioic acid;
octadecenodiacyl carnitines;
taurodeoxycholic acid-3-sulfate;
ceramide;
n, N-trimethyl-5-aminopentanoic acid;
3-hydroxydecanoyl carnitine; and/or
3-hydroxyoctanoyl carnitine.
Example 2
In this example, using the blood sample test data in example 1, all samples were used as training sets, a liver cancer risk prediction model was established using logistic regression, and 100 resampling (boottrap) internal verification was performed. Obtaining a logistic regression model with good liver cancer risk prediction effect.
The present embodiment provides a set of serum metabolic marker combinations for liver cancer risk prediction, comprising:
cystine;
lactate;
oleoylethanolamide;
niu Huangdan cholic acid-3-sulfate;
octadecenodiacyl carnitines;
taurodeoxycholic acid-3-sulfate;
ceramide; and
n, N, N-trimethyl-5-aminopentanoic acid.
The logistic model established by the combination of the 8 serum metabolic markers in this example is y= -0.983 log (cystine abundance) -2.951 log (lactate abundance) +0.878 log (oleoylethanolamine abundance) -0.568 log (Niu Huangdan cholic acid-3-sulfate abundance) +1.163 log (octadecanediol carnitine abundance) +1.361 log (taurodeoxycholic acid-3-sulfate abundance) +1.189 log (ceramide carnitine abundance) +0.967 log (N, N-trimethyl-5-aminopentanoic acid abundance) +1.256.
When the value of y > is-0.43, it is determined that the subject is at high risk of progressing to liver cancer.
The predictive effect of the model is analyzed based on subject work characteristics (Receiver operating characteristic, ROC). In FIG. 2A is the result of a logistic regression model based on the above 8 metabolites, with an area under ROC curve (AUC) of the training set of 0.868 (95% CI: 0.804-0.932), sensitivity and specificity of 87.5% and 72.1%, respectively, and a resampling internal validation calculation with a C index of 0.870 and a Brier score of 0.145.
The method for predicting the liver cancer risk by using the serum metabolic marker for predicting the liver cancer risk comprises the following steps:
1. extracting a serum sample;
a serum sample from the subject was prepared by the method described in "extraction of serum sample" in example 1.
2. Serum metabolic marker detection
Cystine in the serum sample in step 2 of this example was measured according to the method of "3, metabonomics test" in example 1; lactate; oleoylethanolamide; niu Huangdan cholic acid-3-sulfate; octadecenodiacyl carnitines; taurodeoxycholic acid-3-sulfate; ceramide; and the content of N, N, N-trimethyl-5-aminopentanoic acid.
3. Data computation
Subject serum log (cystine abundance) is 0.36, log (lactate abundance) is-0.06, log (oleoylethanolamide abundance) is-0.68, log (Niu Huangdan cholic acid-3-sulfate abundance) is-0.21, log (octadecendioyl carnitine abundance) is-0.44, log (taurodeoxycholic acid-3-sulfate abundance) is 0.55, log (neuroyl carnitine abundance) is-0.97, log (N, N-trimethyl-5-aminopentanoic acid abundance) is 0.96, and the calculated values of the logistic regression model are substituted to be 0.61, greater than-0.43, and the patient is predicted to have a high risk of progressing to cancer.
Example 3
This example uses the blood sample test data of example 1, at a rate of 7: and 3, dividing the liver cancer risk prediction model into a training set and a verification set, and establishing a liver cancer risk prediction model by using a gradient lifting classification tree in machine learning to obtain a gradient lifting tree model with good liver cancer risk prediction effect.
The present embodiment provides a set of serum metabolic marker combinations for liver cancer risk prediction, comprising:
cystine;
oleoylethanolamide;
3-hydroxy-2-propionic acid ethyl ester;
hexadecanedioic acid;
taurodeoxycholic acid-3-sulfate;
ceramide;
3-hydroxydecanoyl carnitine.
In this embodiment, the gradient lifting tree model established by using the combination of the 7 serum metabolic markers is used, the iteration number is 75, the number of randomly extracted features is 100% when the tree is established, the minimum weight used when the tree is lifted is 1, the learning rate is 0.01, the minimum loss required when a leaf partition is newly added is 0.5, the sub-sample data occupies 50% of the whole observed proportion, and the maximum depth of the tree is 2. Substituting the detected abundance of each substance into a gradient lifting tree model calculated value, and judging that the subject has high risk of developing liver cancer when the value is more than 0.51.
The predictive effect of the above model was analyzed based on ROC. B in fig. 2 is the result of the gradient-lifted tree model built based on the 7 metabolites above. The area under ROC curve (AUC) of the training set was 0.911 (95% CI: 0.827-0.963), the predictive probability limit was 0.51, the sensitivity and specificity were 80.5% and 87.5%, respectively, and the C index was 0.911 and the Brier score was 0.197 were calculated directly using the mers2 function in the Hmisc package. In the validation set AUC was 0.930 (95% CI: 0.793-0.988), sensitivity and specificity were 86.7% and 81.0%, respectively, with a calculated C index of 0.930 and Brier score of 0.200 using the threshold of 0.51, the mers2 function.
The method for predicting the liver cancer risk by using the serum metabolic marker for predicting the liver cancer risk comprises the following steps:
1. extracting a serum sample;
a serum sample from the subject was prepared by the method described in "extraction of serum sample" in example 1.
2. Serum metabolic marker detection
Cystine in the serum sample in step 2 of this example was measured according to the method of "3, metabonomics test" in example 1; oleoylethanolamide; 3-hydroxy-2-propionic acid ethyl ester; hexadecanedioic acid; taurodeoxycholic acid-3-sulfate; ceramide; and detecting the content of the 3-hydroxydecanoyl carnitine.
3. Data computation
Subject serum log (cystine abundance) is 0.36, log (oleoylethanolamide abundance) is-0.68, log (3-hydroxy-2-ethyl propionate abundance) is 0.50, log (hexadecanedioic acid abundance) is-0.66, log (taurodeoxycholic acid-3-sulfate abundance) is 0.55, log (ceramide-carnitine abundance) is-0.96, log (3-hydroxydecanoyl-carnitine abundance) is 0.11, and the calculated value of the substituted gradient lifting tree model is 0.58, which is greater than 0.51, and the patient is predicted to have high risk of developing cancer metastasis.
It is apparent that the above examples are given by way of illustration only and are not limiting of the embodiments. Other variations or modifications of the above teachings will be apparent to those of ordinary skill in the art. It is not necessary here nor is it exhaustive of all embodiments. While still being apparent from variations or modifications that may be made by those skilled in the art are within the scope of the invention.
Claims (8)
1. A serum metabolic marker for liver cancer risk prediction, characterized in that the serum metabolic marker comprises the following compounds:
cystine;
lactate;
oleoylethanolamide;
niu Huangdan cholic acid-3-sulfate;
octadecenodiacyl carnitines;
taurodeoxycholic acid-3-sulfate;
ceramide;
n, N, N-trimethyl-5-aminopentanoic acid.
2. A serum metabolic marker for liver cancer risk prediction, characterized in that the serum metabolic marker comprises the following compounds:
cystine;
oleoylethanolamide;
3-hydroxy-2-propionic acid ethyl ester;
hexadecanedioic acid;
taurodeoxycholic acid-3-sulfate;
ceramide;
3-hydroxydecanoyl carnitine.
3. Use of a product for detecting the serum metabolic marker for liver cancer risk prediction according to claim 1 or 2 in the preparation of a product for liver cancer risk prediction.
4. The use according to claim 3, wherein the product for detecting serum metabolic markers for liver cancer risk prediction and/or the product for liver cancer risk prediction comprises a reagent, a test paper, a kit or an instrument.
5. A method for screening serum metabolic markers for liver cancer risk prediction according to claim 1 or 2, comprising the steps of:
step one: collecting a blood sample of a subject at risk of developing liver cancer, and dividing the blood sample into a cancer-transferring group and a non-cancer-transferring group according to whether the subject progresses to liver cancer;
step two: determining the metabolite content in each blood sample, and analyzing the relativity of the determined metabolite content with whether the liver cancer is progressed or not by using more than 3 statistical methods; setting a correlation threshold of each statistical method;
and thirdly, screening out metabolites at least reaching a correlation threshold of one statistical method.
6. The method of claim 5, wherein the statistical method comprises: variable importance projection and difference multiple; the Boruta method; lasso method; extraTree method; or the SelectKBest method;
the correlation threshold value of each statistical method is respectively that the variable importance projection is >1 and the difference multiple is >1.5; ten random tests are carried out by adopting a Borata method, and the first 20 is sorted according to the feature selection times; performing ten random experiments by using Lasso, and sequencing the first 20 according to the feature selection times; performing ten random tests by adopting ExtraTree, and sequencing the first 20 according to the feature selection times; ten random trials were performed using a SelectKBest, ranked 20 before feature selection.
7. The method for screening serum metabolic markers for liver cancer risk prediction according to claim 5, wherein the blood sample of the subject at risk of developing liver cancer is a blood sample of a patient who has hepatitis b or cirrhosis and has not progressed to liver cancer.
8. The method for screening serum metabolic markers for liver cancer risk prediction according to claim 5, wherein in the second step, the metabolite content in each blood sample is measured by means of liquid chromatography-mass spectrometry.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202310451383.3A CN116183924B (en) | 2023-04-25 | 2023-04-25 | Serum metabolism marker for liver cancer risk prediction and screening method and application thereof |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202310451383.3A CN116183924B (en) | 2023-04-25 | 2023-04-25 | Serum metabolism marker for liver cancer risk prediction and screening method and application thereof |
Publications (2)
| Publication Number | Publication Date |
|---|---|
| CN116183924A CN116183924A (en) | 2023-05-30 |
| CN116183924B true CN116183924B (en) | 2023-07-07 |
Family
ID=86438723
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN202310451383.3A Active CN116183924B (en) | 2023-04-25 | 2023-04-25 | Serum metabolism marker for liver cancer risk prediction and screening method and application thereof |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN116183924B (en) |
Family Cites Families (1)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| US20160209428A1 (en) * | 2013-08-21 | 2016-07-21 | The Regents Of The University Of California | Diagnostic and predictive metabolite patterns for disorders affecting the brain and nervous system |
-
2023
- 2023-04-25 CN CN202310451383.3A patent/CN116183924B/en active Active
Also Published As
| Publication number | Publication date |
|---|---|
| CN116183924A (en) | 2023-05-30 |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| CN110057955B (en) | Method for screening specific serum marker of hepatitis B | |
| Brown et al. | Cancer detection and biopsy classification using concurrent histopathological and metabolomic analysis of core biopsies | |
| CN111289736A (en) | Slow obstructive pulmonary early diagnosis marker based on metabonomics and application thereof | |
| US20050101023A1 (en) | Methods for diagnosing urinary tract and prostatic disorders | |
| EP2005172A2 (en) | Apolipoprotein fingerprinting technique | |
| CN110057954B (en) | Application of plasma metabolism marker in diagnosis or monitoring of HBV | |
| CN113960235B (en) | Application and method of biomarker in preparation of lung cancer detection reagent | |
| CN107110867A (en) | Biomarker for diagnosing cancer of liver and application thereof | |
| CN111562338A (en) | Application of transparent renal cell carcinoma metabolic marker in renal cell carcinoma early screening and diagnosis product | |
| Long et al. | Pattern-based diagnosis and screening of differentially expressed serum proteins for rheumatoid arthritis by proteomic fingerprinting | |
| WO2006129401A1 (en) | Screening method for specific protein in proteome comprehensive analysis | |
| Meyer et al. | High‐performance proteomics as a tool in biomarker discovery | |
| Roy et al. | Quantification of proteins and metabolites by mass spectrometry without isotopic labeling | |
| CN101246176B (en) | Mass spectrum kit for detecting squamous-cell carcinoma antigen feminine cervical carcinoma serum protein and preparation method thereof | |
| Song et al. | MALDI‐TOF‐MS analysis in low molecular weight serum peptidome biomarkers for NSCLC | |
| CN116183924B (en) | Serum metabolism marker for liver cancer risk prediction and screening method and application thereof | |
| CN116148482A (en) | Device for breast cancer patient identification and its preparation and use | |
| JP2014032046A (en) | Marker for identification of diabetic renal diseases, and use thereof | |
| CN116165385B (en) | Serum metabolic marker for liver cancer diagnosis and screening method and application thereof | |
| CA2605667A1 (en) | Methods for detecting alzheimer's disease | |
| CN114264828A (en) | Biomarker for identifying benign thyroid nodule and thyroid cancer and application thereof | |
| CN111751551A (en) | Protein molecule as biomarker for diagnosing liver cirrhosis and prognosis method thereof | |
| CN116087482B (en) | Biomarkers for severity typing of course of patients with 2019 novel coronavirus infection | |
| CN117388495B (en) | Application of metabolic marker for diagnosing lung cancer stage and kit | |
| Xiang et al. | Peptidomics: identification of pathogenic and marker peptides |
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 |