CN102323328A - Decision tree model for early diagnosis of lung cancer - Google Patents
Decision tree model for early diagnosis of lung cancer Download PDFInfo
- Publication number
- CN102323328A CN102323328A CN201010561691A CN201010561691A CN102323328A CN 102323328 A CN102323328 A CN 102323328A CN 201010561691 A CN201010561691 A CN 201010561691A CN 201010561691 A CN201010561691 A CN 201010561691A CN 102323328 A CN102323328 A CN 102323328A
- Authority
- CN
- China
- Prior art keywords
- lung cancer
- protein
- chip
- decision tree
- tree model
- 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.)
- Pending
Links
Abstract
The invention relates to the technical field of medical diagnostic, and discloses a decision tree model for early diagnosis of lung cancer, which is characterized in that: a protein chip and time-of-flight mass spectrometry system is used to detect the peripheral serum samples of early lung cancer patients so as to find a specific protein peak obviously different from that of lung cancer patients, and a decision tress model is obtained by a decision tree algorithm according to the mass to charge ratio (m / z) of each protein peak. The decision tree model can be preliminarily used for diagnosis of the lung cancer only by analyzing the m / z of corresponding protein in serum of the detected person and the model provided by the invention. The prediction accuracy rate is 79%.
Description
Technical field
The invention belongs to biological technical field, is a kind of decision-tree model that can be used for diagnosing the early stage of lung cancer, for the early diagnosis of lung cancer provides experiment basis.
Background technology
Lung cancer betides the tunica mucosa bronchiorum epithelium, also claims bronchiolar carcinoma.Many countries report that all the incidence of disease of lung cancer obviously increases over nearly 50 years, and in male sex's cancer knurl patient, lung cancer ranks first, and women's incidence of disease also increases rapidly, account for the 2nd or the 3rd of women's common cancer.Lung cancer originates from the tunica mucosa bronchiorum epithelium, cancerous swelling can be in bronchial lumen or (with) contiguous lung tissue growth, also can pass through lymph, blood is capable or shift through bronchus and to spread.Biological characteristicses such as the histological type of the speed of growth of cancerous swelling and situation that shifts diffusion and cancerous swelling, differentiation degree have certain relation.
The gold instrument that always is considered to proteomics research--two-dimensional polyacrylamide gel electrophoresis technology (2D-PAGE) institute can detected protein also be the sub-fraction in the protein group; Many kinds of protein especially LMWP (MW＜20kD) can not separate out with the method need carry out proteomics research with advanced person's technological means.And a kind of new proteomics research method that just grow up recent years--the surface adds light laser and resolves electrode flight time mass spectrum technology and have special solid phase chromatography surface and can protein be kept above that; After ionization, it is detected, directly obtain proteomic map from untreated biological sample with time of-flight mass spectrometer.Utilize through the solid support of special processing and process protein-chip; According to the difference of protein physics, chemical property, can combine range protein non-specific or specifically, and in mass spectrometer, receive laser bombardment; Various conjugated proteins can be excited and form charged ion; Time length according to the different electric ion flies in electric field forms strong and weak different mass spectra peaks, and then the formation collection of illustrative plates is used for analyzing.This technology is compared with 2D-PAGE, and is simple to operate, highly sensitive, and the sample size that detection needs is few.In recent years should technology use widely having obtained aspect differential expression protein group, protein interaction and the disease detection, be particularly suitable for polypeptide and low molecular weight protein analysis of spectrum.
This technology is obtaining bigger progress aspect the diagnostic marker of seeking diseases such as prostate, mammary gland, ovarian neoplasm.The present invention has set up a kind of model that can be used for diagnosing the early stage of lung cancer from the angle of decision tree, for the early diagnosis of lung cancer provides experiment basis.
Summary of the invention
The present invention utilizes the protein-chip time of-flight mass spectrometer system of Ciphergen company; Detect the protein spectrum of the early stage of lung cancer and healthy subjects serum; Utilize the software analysis early stage of lung cancer control protein group different with normal human serum; Find new early stage of lung cancer correlativity protein, be made into decision-tree model, be used to diagnose the early stage of lung cancer.
The present invention utilizes Ciphergen company protein-chip to detect respectively the early stage of lung cancer and healthy subjects serum; Obtain data and utilize the Biomarker wizard function of the Biosystems software of the said firm; Protein wave spectrum to the early stage of lung cancer and normal group sample compares respectively, draws special protein profiling, finds that according to statistical study the early stage of lung cancer and normal group have the protein peak of statistically-significant difference again; And use the Matlab 2009a software of MathWorks company that the early stage of lung cancer and control group significant difference specific protein are analyzed; Choose 4 mass spectra peaks and (be respectively m/z4505, m/z5689, m/z7580; M/z11245) accuracy is the highest the time, is 79%.Utilize these 4 mass spectra peaks to set up the early stage of lung cancer and normal person's decision-tree model, utilize this model, the mass-to-charge ratio of being examined corresponding proteins matter in the human serum and model of the present invention are compared one by one, can be early stage of lung cancer diagnosis auxiliary guidance is provided.
The detection method of decision-tree model is following:
1. serum is handled
Put into 4 ℃ of refrigerators preservations immediately after gathering preceding blood of early stage of lung cancer patient art and normal control peripheric venous blood, subsequent use.During detection, melt freezing serum, centrifugal, get blood serum sample with damping fluid, jolting.
2. the balance of chip
Add SAS in the chip hole, balance chip 2 times, each 5min.
3. dilution
Get blood serum sample, add urea, vibration 30min adds in the SAS, the mixing dilution.
4. go up appearance
The blood serum sample of getting after the dilution adds the good chip well of balance, puts chip and hatches 90min in shaking table.
5. sample wash-out
Get SAS and add the chip well, wash-out 2 times, each 5min.Remove ionized water flushing chip.
6. chip and protein bound reaction
Get energy absorber and add the chip well, treat to repeat again to add absorbing agent once after the nature volatilization.
7. chip detection
Chip is placed the protein chip reading machine, rectify an instrument with the chip that is added with standard protein.Chip reading apparatus parameter is set, collects data with CipheEgen Pmteinchip 3.1.1 software.Wherein ordinate is the protein relative content, and horizontal ordinate is protein mass-to-charge ratio (M/Z).Through clinic trial, this decision-tree model can be distinguished the early stage of lung cancer and normal control preferably, and this decision tree accuracy is 79%.
Description of drawings
The decision-tree model that Fig. 1 lung cancer and normal person differentiate
M/z representes the mass-to-charge ratio of differential protein among the figure, and case represents early stage of lung cancer patient, and ctrl represents the normal person.
Embodiment
1. material
1.1 detection serum
Get experimenter's venous blood 5ml, centrifugal after 4 degree leave standstill 2 hours in the refrigerators (4000rpm, 5min). get serum, put low temperature refrigerator (negative 80 degree) or liquid nitrogen and preserve, subsequent use.
1.2 required main agents and instrument
Sodium acetate, acetonitrile, trifluoroacetic acid, sinapinic acid (SPA) are all available from Shanghai chemical reagents corporation.CMlO protein chip and protein-chip reading machine (PBS II type) are available from Ciphergen company.
2. method
2.1 the balance of chip: sodium acetate (NaAc) solution of getting 200ul 50mmol/L pH 4.0 adds in the well of chip, balance chip 2 times, each 5min.
2.2 dilution
Get the 6ul blood serum sample, add 12ul 9mmol/L urea, vibration 30min adds 222ul 50mmol/LpH and is in 3.0 the NaAc solution, the mixing dilution.
2.3. last appearance
The blood serum sample 200ul/ hole of getting after the dilution adds the good chip well of balance, puts chip and hatches 90min in shaking table.
2.4. sample wash-out
The NaAc solution 200ul/ hole of getting 50mmol/L pH 4.0 adds the chip well, wash-out 2 times, each 5min.Get the deionized water rinsing chip in 300ul/ hole.
2.5. chip and protein bound reaction
Get energy absorber SPA (containing saturated sinapinic acid solution and massfraction 50% acetonitrile and 0.15% trifluoroacetic acid) 0.5ul/ hole and add the chip well, treat to repeat again to add SPA once after the nature volatilization.
2.6. chip detection
Chip is placed the protein chip reading machine, rectify an instrument to 0.1% scope with the NP20 chip that is added with the AU-in_one standard protein.Chip reading apparatus parameter is provided with: laser intensity 150, and detection sensitivity 7, optimization range 1984.2～9921u, highest weight 49605u, each point on the chip is gathered 130 times.Collect data with CipheEgen Proteinchip 3.1.1 software.Wherein ordinate is the protein relative content, and horizontal ordinate is protein mass-to-charge ratio (M/Z).The person under inspection's that analysis is obtained haemocyanin wave spectrum compares according to specific protein mass-to-charge ratio in the decision-tree model (Fig. 1), whether suffers from the early stage of lung cancer (seeing embodiment) to judge the person under inspection.
Experiment embodiment
Get person under inspection's serum, obtain its haemocyanin collection of illustrative plates, press the decision-tree model that Fig. 1 early stage of lung cancer and normal group are differentiated through check; The observing protein crest sees that earlier mass-to-charge ratio M/Z is 4505 protein, and≤3.16 whether its peak intensity coefficient A; Be to see again that then M/Z is 5689 protein; ≤1.18 whether its peak intensity coefficient A be then to be control group, otherwise be the case group.See that mass-to-charge ratio M/Z is 4505 protein; ≤0.14 whether its peak intensity coefficient A＞saw again that M/Z was 11245 protein at 3.16 o'clock, its peak intensity coefficient A are then to be control group; Otherwise see that at last M/Z is 7580 protein; ≤31.25 whether its peak intensity coefficient A be then to be the case group, otherwise be control group.
More than be the description of this invention and non-limiting, based on other embodiment of inventive concept, all among protection scope of the present invention.
Claims (1)
1. the decision-tree model that is used for the lung cancer early diagnosis; The serum proteins collection of illustrative plates that it is characterized in that being detected by protein-chip time of-flight mass spectrometer system analysis obtains a plurality of specific proteins of patients with lung cancer through software statistics mass-to-charge ratio (m/z) and protein peak strength factor A thereof draw and form; The mass-to-charge ratio of said specific protein (m/z) is respectively m/z4505; M/z5689, m/z7580, m/z11245.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201010561691A CN102323328A (en) | 2010-11-25 | 2010-11-25 | Decision tree model for early diagnosis of lung cancer |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201010561691A CN102323328A (en) | 2010-11-25 | 2010-11-25 | Decision tree model for early diagnosis of lung cancer |
Publications (1)
| Publication Number | Publication Date |
|---|---|
| CN102323328A true CN102323328A (en) | 2012-01-18 |
Family
ID=45451112
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN201010561691A Pending CN102323328A (en) | 2010-11-25 | 2010-11-25 | Decision tree model for early diagnosis of lung cancer |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN102323328A (en) |
Cited By (2)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN106897570A (en) * | 2017-03-02 | 2017-06-27 | 山东师范大学 | A kind of COPD test system based on machine learning |
| CN111582489A (en) * | 2020-05-14 | 2020-08-25 | 上海深至信息科技有限公司 | Distributed deployment decision making system and method of ultrasonic artificial intelligence model |
-
2010
- 2010-11-25 CN CN201010561691A patent/CN102323328A/en active Pending
Cited By (4)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN106897570A (en) * | 2017-03-02 | 2017-06-27 | 山东师范大学 | A kind of COPD test system based on machine learning |
| CN106897570B (en) * | 2017-03-02 | 2021-05-11 | 山东师范大学 | Chronic obstructive pulmonary disease testing system based on machine learning |
| CN111582489A (en) * | 2020-05-14 | 2020-08-25 | 上海深至信息科技有限公司 | Distributed deployment decision making system and method of ultrasonic artificial intelligence model |
| CN111582489B (en) * | 2020-05-14 | 2023-07-14 | 上海深至信息科技有限公司 | Distributed deployment decision-making system and method of ultrasonic artificial intelligent model |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| CN109884302B (en) | Lung cancer early diagnosis marker based on metabonomics and artificial intelligence technology and application thereof | |
| CN105277718B (en) | For the product of the examination of malignant tumour correlation and assessment, application and method | |
| CN102565318B (en) | Reagent for liver cancer monitoring, staging and prognosis risk assessment and method thereof | |
| Hu et al. | Prediction of gastric cancer metastasis through urinary metabolomic investigation using GC/MS | |
| CN106950379B (en) | A kind of lung cancer monitoring reagent box and its application method | |
| Ye et al. | A vision for better health: mass spectrometry imaging for clinical diagnostics | |
| RU2015119512A (en) | METHYLGIOXAL AS A MALIGNANT TUMOR MARKER | |
| CN112151121A (en) | Diagnostic marker, kit and screening method for esophageal cancer diagnosis and construction method of esophageal cancer diagnosis model | |
| CN106370744A (en) | Application of 8-hydroxydeoxyguanosine as urine marker | |
| WO2013162773A9 (en) | Quantitation of biomarkers for the detection of prostate cancer | |
| CN101451975B (en) | Method for detecting cancer of stomach prognosis and staging blood serum protein | |
| CN102323328A (en) | Decision tree model for early diagnosis of lung cancer | |
| Quanico et al. | Combined MALDI mass spectrometry imaging and parafilm-assisted microdissection-based LC-MS/MS workflows in the study of the brain | |
| CN102200537A (en) | Protein fingerprint mass spectrum model for lung cancer early diagnosis | |
| CN112986441A (en) | Tumor marker screened from tissue metabolism contour, application thereof and auxiliary diagnosis method | |
| US20140162903A1 (en) | Metabolite Biomarkers For Forecasting The Outcome of Preoperative Chemotherapy For Breast Cancer Treatment | |
| CN101246176B (en) | Mass spectrum kit for detecting squamous-cell carcinoma antigen feminine cervical carcinoma serum protein and preparation method thereof | |
| CN113514530A (en) | Thyroid malignant tumor diagnosis system based on open ion source | |
| CN107677756A (en) | Detect the method and its application of dissection of aorta peripheral blood small molecule metabolic markers | |
| CN102323324A (en) | A kind of decision-tree model of early stage tumor of breast diagnosis | |
| CN1707258A (en) | Method for early detecting gastric cancer from blood serum | |
| CN101329346A (en) | Optimizing mass spectrogram model for detecting breast cancer characteristic protein and preparation method and application thereof | |
| CN102207506A (en) | Decision tree model for diagnosis of early-stage prostatic cancer | |
| CN102324001A (en) | Method for predicting gastric cancer on the basis of high performance liquid chromatography/mass spectrometry (HPLC/MS) metabonomics data analysis | |
| CN113687076B (en) | Combined detection serum marker for early diagnosis of lung adenocarcinoma and application thereof |
Legal Events
| Date | Code | Title | Description |
|---|---|---|---|
| C06 | Publication | ||
| PB01 | Publication | ||
| C10 | Entry into substantive examination | ||
| SE01 | Entry into force of request for substantive examination | ||
| C02 | Deemed withdrawal of patent application after publication (patent law 2001) | ||
| WD01 | Invention patent application deemed withdrawn after publication |
Application publication date: 20120118 |