KR20140057688A - Apparatus for diagnosis colorectal cancer - Google Patents

Abstract

The present invention provides a colorectal cancer diagnostic device for confirming low mass ion to diagnose colorectal cancer through biostatics analysis of low mass ion extracted from a biological sample, and for diagnosing colorectal cancer by using a mass spectrum of the low mass ion. The present invention provides a discriminant for discriminating a colorectal cancer patient group from a non-patient group. In other words, the present invention provides a discriminant for showing an excellent discriminative performance for not only a training group for drawing a discriminant, but also an independent verification group, and the colorectal cancer diagnostic device for diagnosing colorectal cancer by confirming the low mass ions.

Description

{Apparatus for diagnosis colorectal cancer}

The present invention relates to an apparatus for diagnosing colorectal cancer, and more particularly, to a mass spectrometry of low mass ions for diagnosis of colorectal cancer through biostatistical analysis of low mass ions extracted from a biological sample, The present invention relates to a colon cancer diagnostic apparatus capable of diagnosing colon cancer using an ion mass spectrum.

Cancer is a disease in which cells infiltrate infinitely and interfere with the function of normal cells. Lung cancer, gastric cancer, breast cancer (BRC), colorectal cancer (CRC) Can occur in any organization. Early diagnosis of cancer was based on external changes of the biotissue due to the growth of cancer cells. Recently, however, diagnosis using biomolecules such as blood, glycocyte, DNA, And detection has been attempted. However, the most commonly used cancer diagnosis method is a tissue sample obtained through biopsy or an image diagnosis.

Among them, biopsy results in great pain for the patient, and it is not only expensive but also takes a long time to diagnose. In addition, when the patient is actually cancerous, there is a risk that cancer metastasis may be induced during the biopsy, and in the case where a tissue sample can not be obtained through biopsy, It is impossible to diagnose the disease until the removal of the disease.

In the diagnosis using imaging, cancer is judged based on X-ray image, nuclear magnetic resonance (NMR) image obtained using a contrast agent having a disease target substance, and the like. However, there is a possibility of misdiagnosis according to the skill of the clinician or the reader, and this image diagnosis has a disadvantage that it depends greatly on the precision of the device for obtaining the image. Furthermore, even the most precise apparatus can not detect tumors of a few mm or less, which is difficult to detect in the early stages of onset. In addition, in order to obtain an image, a patient or a disease-capable person is exposed to electromagnetic waves of high energy capable of causing mutation of a gene, so that it may cause another disease, and there is a disadvantage that there are limitations on the number of diagnosis through imaging .

In the case of the digestive system, the presence or absence of the disease is usually judged by visual observation using an endoscope. However, even if the procedure is very painful to the patient and an abnormality is found through visual observation, malignant / benign tumors, Biopsy should be performed to determine the exact disease.

In particular, CRC is a common cancer with a worldwide incidence rate of less than 3, and the possibility of treatment is highly dependent on the stage of cancer. In other words, it has a very high cure rate when it is detected at early stage through early diagnosis. Therefore, it is common to recognize the disease by the color change of the feces due to hemorrhage, and the patient or the person having the disease can be examined even though the diagnosis of the disease is important and the abnormality accompanying the cancer progresses is small. Observation through colonoscopy is common and biopsy should be performed for accurate disease identification. In summary, early diagnosis of CRC is important, and colonoscopy and biopsy are time-consuming, costly, inconvenient, and painful. Therefore, diagnostic methods that can dramatically reduce the number of unnecessary colonoscopy and biopsy specimens Is required.

Therefore, it would be very useful for patients if a new molecular approach could be used to screen CRC at an early stage. Genomics, proteomics, and molecular pathology have provided a number of biomarker candidates with clinically significant potential. Although it is thought that the therapeutic effect can be improved by actively using them for the stage of cancer and personalized treatment for each patient, many studies should be preceded in order to be applied to clinical treatment.

Recent CRC screening tests are performed using a colonoscopy to determine whether a gross abnormality is present or to perform a fecal occult blood test (FOBT). Colonoscopy has been used as a standard method in CRC screening tests, but it is limited to invasive and acceptable patients. Thus, much effort has recently been focused on fecal tests, which are noninvasive, do not require intestinal lavage, and allow delivery of specimens. Fecal markers can be classified as leaking, secreted, or exfoliated from a tumor. For example, in traditional FOBT, hemoglobin has been recognized as a breeding marker in a large-scale screening program to diagnose CRC, but its ability to discriminate markers to date, including this, is unsatisfactory.

On the other hand, the mass spectrometer of MALDI-TOF (mass-assisted laser desorption / ionization-time of flight) mass spectrometer can extract the spectrum of mass ion in the blood. The mass spectrometry used in the existing protein studies mainly focused on the mass value range of 800 to 2500 m / z, which is the mass value range of the peptide when the protein is truncated with trypsin to be. In addition, mass spectra of low-mass ions can be extracted using a MALDI-TOF mass spectrometer. However, since the low mass band below about 800 m / z is a region where the mass ions of the matrix are mixed rather than the object of analysis, research on this region has not been actively carried out.

The mass spectrum of the extracted low mass ions can be analyzed by MarkerView ^TM (version version 1.2, abbreviated below) which is one of conventional software. The present inventors analyzed the mass spectrum of low mass ions extracted from the serum of the CRC patient group and the normal control (control, CONT) using MarkerView ^TM , and the method will be described in detail with reference to FIG.

The low mass ion mass spectrum of the T2D file format, extracted from the serum collected from a set of 133 CRC patients (Table 1) and 153 normal control group (Set A ₁ ) of Table 2 using a MALDI-TOF mass spectrometer, ^And imported into the ^TM (A11).

CRC Sex Age
year Stage Location Cell Type CEA
ng / mL CRC-A1 M 77 I A-colon AC 1.8 CRC-A2 M 50 I Rectum AC 1.9 CRC-A3 F 47 I S-colon AC 0.7 CRC-A4 F 56 III S-colon AC 1.2 CRC-A5 F 82 I A-colon AC 1.1 CRC-A6 M 59 I Rectum AC 1.9 CRC-A7 M 73 I Rectum AC 3.6 CRC-A8 M 71 I S-colon AC 3.6 CRC-A9 M 50 I S-colon AC 2.5 CRC-A10 M 56 I S-colon AC 7.3 CRC-A11 M 61 I Rectum AC 7.7 CRC-A12 F 78 I Rectum AC 2.6 CRC-A13 M 64 I S-colon AC 1.8 CRC-A14 F 50 I Rectum AC 1.6 CRC-A15 F 59 I Rectum AC 1.6 CRC-A16 M 73 I Rectum AC 1.9 CRC-A17 M 65 I S-colon AC 14.0 CRC-A18 M 72 I S-colon AC 4.6 CRC-A19 M 82 I Rectum AC 3.2 CRC-A20 M 52 III S-colon AC 3.2 CRC-A21 F 59 III S-colon AC 1.7 CRC-A22 F 73 III S-colon AC 5.7 CRC-A23 M 70 III S-colon AC 3.6 CRC-A24 M 75 II A-colon AC 2.1 CRC-A25 F 81 II S-colon AC 4.1 CRC-A26 F 76 II Rectum AC 25.3 CRC-A27 F 71 II A-colon AC 1.6 CRC-A28 M 72 II A-colon AC 3.8 CRC-A29 F 82 II S-colon AC 1.8 CRC-A30 F 68 II D-colon AC 1.7 CRC-A31 M 71 II S-colon AC 3.6 CRC-A32 F 67 II A-colon AC 1.9 CRC-A33 M 45 II D-colon MAC 3.3 CRC-A34 M 60 II S-colon AC 2.8 CRC-A35 M 74 II S-colon AC 5.3 CRC-A36 M 57 II Rectum AC 7.3 CRC-A37 F 65 II S-colon AC 2.1 CRC-A38 M 77 II A-colon AC 1.5 CRC-A39 M 71 II D-colon AC 4.1 CRC-A40 F 66 II Rectum AC 4.3 CRC-A41 F 49 II A-colon AC 1.6 CRC-A42 F 79 II A-colon AC 2.9 CRC-A43 M 69 II S-colon AC 4.2 CRC-A44 M 66 II S-colon AC 12.0 CRC-A45 M 74 II A-colon AC 1.5 CRC-A46 M 69 II T-colon AC 1.2 CRC-A47 M 43 II S-colon AC 2.2 CRC-A48 F 67 II A-colon AC 1.4 CRC-A49 M 72 II A-colon AC 4.9 CRC-A50 F 67 II A-colon AC 7.3 CRC-A51 F 75 II Rectum AC 12.6 CRC-A52 M 68 II D-colon AC 4.7 CRC-A53 F 60 II S-colon AC 3.3 CRC-A54 M 74 II S-colon AC 9.0 CRC-A55 M 68 III A-colon AC 9.2 CRC-A56 F 55 III Rectum AC 2.1 CRC-A57 F 61 III A-colon AC 12.7 CRC-A58 M 59 III S-colon AC 2.7 CRC-A59 M 67 III Rectum AC 9.5 CRC-A60 M 48 III S-colon AC 1.3 CRC-A61 M 58 III Rectum AC 1.7 CRC-A62 F 50 III S-colon AC 4.8 CRC-A63 F 51 III S-colon AC 7.0 CRC-A64 F 74 III T-colon AC 2.5 CRC-A65 M 60 III Rectum AC 3.5 CRC-A66 M 52 III S-colon AC 2.5 CRC-A67 M 54 III A-colon AC 5.3 CRC-A68 M 82 III S-colon AC 2.4 CRC-A69 M 54 III S-colon AC 5.3 CRC-A70 F 79 III Rectum AC 14.1 CRC-A71 F 44 III S-colon AC 1.4 CRC-A72 M 66 III Rectum AC 1.2 CRC-A73 M 53 III A-colon AC 4.2 CRC-A74 M 64 III T-colon AC 1.8 CRC-A75 F 42 III S-colon AC 0.8 CRC-A76 M 49 III Rectum AC 2.7 CRC-A77 M 68 III Rectum AC 3.9 CRC-A78 M 51 III S-colon AC 5.2 CRC-A79 M 64 III Rectum AC 7.7 CRC-A80 M 42 III S-colon AC 2.8 CRC-A81 F 43 III A-colon AC 4.7 CRC-A82 M 66 III S-colon AC 9.1 CRC-A83 M 37 III Rectum AC 3.7 CRC-A84 F 81 III Rectum AC 8.4 CRC-A85 F 73 III S-colon AC 1.7 CRC-A86 M 54 III Rectum AC 6.4 CRC-A87 F 58 III Rectum AC 21.3 CRC-A88 F 42 III Rectum AC 0.7 CRC-A89 F 50 III D-colon AC 6.4 CRC-A90 M 56 III S-colon AC 7.3 CRC-A91 F 58 III S-colon AC 2.1 CRC-A92 F 70 IV Rectum AC 3.9 CRC-A93 M 68 IV Rectum AC 6.0 CRC-A94 M 53 IV Rectum AC 54.7 CRC-A95 F 63 IV D-colon AC 12.3 CRC-A96 F 63 IV A-colon AC 1.4 CRC-A97 M 63 II D-colon AC 4.9 CRC-A98 F 66 II S-colon AC 4.2 CRC-A99 M 48 II Rectum AC 28.4 CRC-A100 M 68 II S-colon AC 2.3 CRC-A101 M 48 II S-colon AC 4.8 CRC-A102 F 81 II S-colon AC 2.4 CRC-A103 M 56 II A-colon AC 34.6 CRC-A104 M 71 III Rectum AC 16.5 CRC-A105 M 66 III S-colon AC 689.8 CRC-A106 M 65 III D-colon AC 3.4 CRC-A107 F 65 III S-colon MAC 2.7 CRC-A108 F 51 III Rectum AC 1.4 CRC-A109 M 58 III S-colon AC 2.8 CRC-A110 F 48 III A-colon AC 0.9 CRC-A111 M 71 III S-colon AC 6.0 CRC-A112 M 68 III A-colon AC 2.7 CRC-A113 F 54 III A-colon AC 1.7 CRC-A114 M 66 IV S-colon AC 6.4 CRC-A115 F 72 IV A-colon AC 73.4 CRC-A116 F 69 IV A-colon AC 49.0 CRC-A117 M 75 IV S-colon AC 16.7 CRC-A118 F 49 III S-colon AC 1.0 CRC-A119 F 63 III A-colon AC 58.2 CRC-A120 M 74 III A-colon AC 2.8 CRC-A121 F 54 III T-colon AC 2.2 CRC-A122 M 68 III Rectum AC 22.5 CRC-A123 M 66 III Rectum MAC 1.2 CRC-A124 M 72 IV Rectum SC 8.2 CRC-A125 M 73 IV Rectum AC 52.2 CRC-A126 M 54 IV A-colon AC 2.0 CRC-A127 F 54 IV S-colon AC 29.8 CRC-A128 M 43 IV Rectum AC 36.4 CRC-A129 F 52 IV A-colon MAC 9.0 CRC-A130 M 48 IV S-colon AC 15.9 CRC-A131 M 62 IV Rectum AC 6.3 CRC-A132 M 77 I D-colon AC 6.4 CRC-A133 M 78 I Rectum AC 2.7

AC: Adenocarcinoma

CEA: Carcinoembryonic antigen

MAC: Mucinous adenocarcinoma

Control Sex Age
year CEA
ng / mL Control Sex Age
year CEA
ng / mL CONT-A1 M 39 1.3 CONT-A78 F 64 2.9 CONT-A2 F 70 1.2 CONT-A79 F 52 1.9 CONT-A3 M 66 1.3 CONT-A80 F 37 2.1 CONT-A4 M 53 0.8 CONT-A81 F 49 2.6 CONT-A5 F 69 1.0 CONT-A82 F 48 1.5 CONT-A6 F 68 1.8 CONT-A83 F 30 <0.5 CONT-A7 M 35 1.7 CONT-A84 F 56 1.4 CONT-A8 M 62 3.7 CONT-A85 F 50 1.2 CONT-A9 M 62 1.1 CONT-A86 F 49 2.1 CONT-A10 M 48 5.3 CONT-A87 F 38 0.6 CONT-A11 M 48 1.8 CONT-A88 F 59 1.6 CONT-A12 M 66 1.6 CONT-A89 F 51 1.0 CONT-A13 M 66 1.4 CONT-A90 F 41 1.8 CONT-A14 M 66 4.2 CONT-A91 F 48 1.2 CONT-A15 M 54 1.4 CONT-A92 F 39 0.5 CONT-A16 M 54 1.0 CONT-A93 F 51 1.1 CONT-A17 M 62 2.0 CONT-A94 F 44 1.5 CONT-A18 F 45 0.7 CONT-A95 F 38 1.5 CONT-A19 M 39 3.2 CONT-A96 F 48 1.9 CONT-A20 M 67 1.8 CONT-A97 F 70 4.8 CONT-A21 M 63 5.5 CONT-A98 F 54 2.8 CONT-A22 M 48 2.8 CONT-A99 F 38 2.8 CONT-A23 M 66 3.2 CONT-A100 F 50 1.1 CONT-A24 M 55 5.0 CONT-A101 F 54 1.8 CONT-A25 M 55 1.0 CONT-A102 M 49 1.2 CONT-A26 M 62 7.0 CONT-A103 F 38 0.9 CONT-A27 F 57 1.2 CONT-A104 F 44 - CONT-A28 M 61 0.9 CONT-A105 M 52 - CONT-A29 M 50 1.9 CONT-A106 F 45 - CONT-A30 M 46 1.5 CONT-A107 F 54 - CONT-A31 M 51 4.0 CONT-A108 F 51 3.1 CONT-A32 F 68 1.8 CONT-A109 M 54 6.4 CONT-A33 F 68 1.4 CONT-A110 M 46 1.1 CONT-A34 M 64 1.7 CONT-A111 M 47 1.8 CONT-A35 M 30 0.7 CONT-A112 M 49 1.7 CONT-A36 F 52 2.1 CONT-A113 F 55 <0.5 CONT-A37 F 59 1.2 CONT-A114 M 36 0.7 CONT-A38 M 53 1.6 CONT-A115 M 59 0.8 CONT-A39 F 69 1.2 CONT-A116 M 46 3.7 CONT-A40 F 68 1.0 CONT-A117 F 46 <0.5 CONT-A41 F 65 3.1 CONT-A118 F 50 0.9 CONT-A42 M 31 1.2 CONT-A119 M 58 - CONT-A43 F 59 0.7 CONT-A120 M 34 1.7 CONT-A44 M 43 1.4 CONT-A121 M 53 2.9 CONT-A45 M 66 2.3 CONT-A122 M 45 3.7 CONT-A46 M 48 4.2 CONT-A123 M 47 4.5 CONT-A47 F 41 2.1 CONT-A124 F 34 0.6 CONT-A48 F 65 3.8 CONT-A125 F 58 1.5 CONT-A49 F 67 1.5 CONT-A126 F 54 - CONT-A50 F 45 0.6 CONT-A127 M 35 1.8 CONT-A51 M 30 1.0 CONT-A128 M 49 1.4 CONT-A52 M 55 1.2 CONT-A129 M 48 3.2 CONT-A53 M 54 2.1 CONT-A130 F 34 <0.5 CONT-A54 M 69 2.8 CONT-A131 M 45 4.4 CONT-A55 M 53 1.8 CONT-A132 F 45 0.8 CONT-A56 F 47 1.7 CONT-A133 M 52 - CONT-A57 M 31 1.7 CONT-A134 F 44 - CONT-A58 M 53 3.2 CONT-A135 F 46 - CONT-A59 F 49 1.4 CONT-A136 M 58 - CONT-A60 M 62 1.7 CONT-A137 M 45 4.3 CONT-A61 M 31 2.3 CONT-A138 M 61 1.4 CONT-A62 M 40 0.8 CONT-A139 M 42 2.7 CONT-A63 F 49 1.4 CONT-A140 M 48 3.0 CONT-A64 F 33 1.7 CONT-A141 M 53 1.9 CONT-A65 M 51 3.4 CONT-A142 F 54 2.3 CONT-A66 M 52 2.0 CONT-A143 F 39 1.3 CONT-A67 F 66 1.3 CONT-A144 F 55 1.3 CONT-A68 M 56 1.9 CONT-A145 M 53 - CONT-A69 F 65 1.4 CONT-A146 F 46 - CONT-A70 M 50 1.4 CONT-A147 F 45 - CONT-A71 M 54 1.3 CONT-A148 F 63 - CONT-A72 M 68 1.6 CONT-A149 F 51 - CONT-A73 M 59 2.5 CONT-A150 M 51 - CONT-A74 F 51 2.1 CONT-A151 F 52 - CONT-A75 F 39 0.8 CONT-A152 F 52 - CONT-A76 F 40 1.5 CONT-A153 F 70 - CONT-A77 F 50 1.9

The conditions of Table 3 were used as import conditions.

Mass tolerance 100 ppm Minimum required response 10.0 Maximum number of peaks 10000

Next, the imported peak intensities were normalized (A12). There are a number of normalization methods in MarkerView ^TM . Normalization is performed using the "Normalization Using Total Area Sums" method. In this method, the subscales of intensity for each sample are obtained, and the average of the subscales for each sample is obtained. Then, each peak intensity is multiplied by a scaling factor for each sample so that the subscale of each sample coincides with the average. That is, after the normalization as described above, the subtotals of the intensities of the respective samples become equal.

Next, the normalized peak intensities were Pareto scaled (A13). That is, the peak intensities of the normalized peak intensities are divided by the square root of the standard deviation, and the peak intensities are Pareto-scaled.

Next, the discriminant score (DS) was calculated by performing principal component analysis-based linear discriminant analysis (PCA-DA) on the Pareto-scaled peak intensities (A14). That is, the principal component analysis is performed in two steps to obtain factor loading, which is a weighting factor for each mass ion. The resultant value is multiplied by the Pareto-scaled intensity, And the number of stars discriminated was calculated. Since the maximum number of peaks was set to 10,000 in the import condition of Table 3 and a sufficient number of samples were imported together, the parameter load value was calculated to 10,000 in this calculation, and therefore, one discrimination score was calculated by adding 10,000 terms.

Next, it is determined whether the calculated discrimination score is a positive number (A15), positive (A16) if negative, negative (A17) if negative. In other words, it was applied to CRC, and it was judged as CRC patients when positive and negative when negative.

FIG. 2 shows the distribution of discrimination scores calculated by the method of FIG. 1 for a group consisting of 133 CRC patients and 153 normal control patients who were already clinically determined. The results of the determination according to the discrimination score shown in FIG. 2 are summarized by using a confusion matrix.

Sensitivity 98.50% Set A ₁ True CRC True CONT Specificity 98.69% Predicted CRC 131 2 PPV 98.50% Predicted CONT 2 151 NPV 98.69%

, Prior art MarkerView ^TM principal component analysis based on linear discriminant through the analysis sensitivity (sensitivity), the specificity of (specificity), positive predictive value (positive predictive value, PPV) As can be seen through the 2 and 4, the audio Very good discrimination results were obtained with a predictive value (NPV) of more than 98%.

However, the robustness of the equation must be verified before it can be used in clinical practice. In other words, we should still show good discrimination results for the mass spectra that were measured once and repeatedly several times in addition to the set that constructed the discriminant, and the new CRC patient group and the normal control group It should be possible to obtain good discrimination results with the same discriminant expression. Repeated measurements of mass spectra may involve the process of freezing and thawing serum, or the process of extracting serum by mixing it with fresh methanol / chloroform. These processes can be regarded as disturbances in the statistical analysis of the mass spectrum, and can only be applied to clinical applications by constructing an insensitive discriminant to these disturbances.

In other words, the conventional principal component analysis based linear discriminant analysis method described with reference to FIGS. 1, 2 and 4 is applied to a set of specific samples individually, that is, in discrimination of individual training sets Although the result is good, the result of the discrimination in the validation set is unsatisfactory.

The reason why the discriminant equation does not have robustness is that the significant number of mass ions among the 10,000 mass ions constituting the discriminant equation are at least unnecessary for the discrimination between the CRC patient group and the normal control group It is presumed that sometimes the discrimination of the training set does not cause a problem, but the discrimination of the verification set contains the mass ions which can potentially confuse the discrimination result.

Therefore, it will be necessary to find out only indispensable mass ions in order to obtain a strong and robust discrimination result through the process of positively removing mass ions which are at least unnecessary or potentially confusing the result of discrimination .

The present invention relates to a method for diagnosing colorectal cancer by identifying a low mass ion mass spectrum for diagnosing colorectal cancer through biostatistical analysis of low mass ions extracted from a biological sample and for diagnosing colorectal cancer using this low mass ion mass spectrum Provided is a colon cancer diagnostic apparatus.

The present invention proposes a discriminant equation that gives robust discrimination results for CRC patients and non-patient samples. The present invention also provides a mass discrimination formula for a CRC patient and a non-patient sample, We propose a discriminant equation with more than 85% sensitivity, specificity, positive predictive value, and negative predictive value for all mass spectrums repeatedly measured for patient aggregation. We can diagnose CRC by checking low mass ions constituting it A colon cancer diagnostic apparatus.

In order to solve the above-mentioned problems, the present invention provides a colon cancer diagnostic apparatus comprising: a low-mass ion detector for detecting a mass spectrum of a low-mass ion from a biological sample of a plurality of colorectal cancer patients and non-patient cases; A colon cancer diagnosis unit for comparing and analyzing the mass spectral patterns of the low-mass ions to determine colon cancer diagnosis information; And a display unit for converting the colorectal cancer diagnosis information determined by the colorectal cancer diagnosis unit into an outputable form and displaying the converted colorimetric diagnosis information.

Wherein the low mass ion detector comprises: Mass spectra of the low mass ions can be extracted by detecting peak intensity of low mass ions from the biological sample. Wherein the low mass ion detector comprises: And a mass spectrometer.

Wherein the colon cancer diagnosis unit comprises: First alignment means for aligning low mass ion mass spectra of said colorectal cancer patients and non-patient cases which are training candidate sets; A first discrimination score calculating means for calculating a discrimination score by performing a biostatistical analysis on the sorted mass spectrum; A factor load calculation means for calculating a sensitivity and a specificity according to the discrimination score and calculating a factor loading by selecting a first training set based on the sensitivity and the specificity, An ion selection means for diagnosing colon cancer which selects low mass ions for diagnosis of colon cancer based on the discrimination performance among candidate low mass ions satisfying the candidate condition; Second alignment means for aligning the low mass ion mass spectrum of the identified biological sample with the first training set; A second discrimination score calculation means for calculating a discrimination score from the peak intensity of the low-mass ion of the discrimination object and the factor load value; And colorectal cancer determination means for determining the discriminant to be positive or negative for colorectal cancer according to the discrimination score,

The colon cancer diagnostic ion selecting means comprises: The plurality of colorectal cancer patients and non-patient cases are classified into first type discrimination cases each comprising a plurality of colon cancer patient cases and a plurality of normal case cases, And the second type discrimination cases are respectively executed for the first type discrimination cases and the second type discrimination cases so that the low mass ions for diagnosing colorectal cancer are respectively classified into the first type discrimination cases and the second type discrimination cases, A low mass ion for diagnosing first type colorectal cancer and a low mass ion for diagnosis of second type colorectal cancer in the second type discriminating case.

The mass values of the low mass ions for diagnosis of the first type colorectal cancer are 18.0260, 22.9797, 74.0948, 76.0763, 102.0916, 105.1078, 106.0899, 107.0477, 118.0822, 123.0395, 137.0423, 137.0729, 147.0573, 147.1058, 169.0653, 181.0656, 190.0849, 191.0848 , 191.3324, 195.0785, 212.3195, 231.0667, 235.0053, 256.0939, 266.9557, 267.9501, 288.2033, 291.0997, 295.0663, 300.1297, 301.1269, 316.2288, 317.2311, 335.1862, 340.2241, 343.2451, 345.2583, 357.0666, 357.2784, 366.2310, 368.2551, 369.3302, 377.0570 , 379.1438, 379.4765, 383.0529, 384.1745, 388.2688, 401.0531, 423.0313, 428.1878, 454.2090, 465.3014, 466.1923, 468.1851, 469.2831, 477.1721, 478.1678, 480.1715, 482.3220, 483.3258, 496.8683, 497.7636, 503.8719, 508.3407, 510.3265, 512.3119, 513.3177 , 518.2931, 518.8555, 519.2967, 519.8598, 525.3449, 534.2739, 537.2800, 538.3306, 540.2629, 540.8144, 542.8457, 544.8692, 548.2856, 566.8375, 581.1957, 582.1888, 583.2242, 656.0270, 667.3291, 709.3519, 710.3581, 711.3617, 712.3683, 713.3798, 991.6196 , 99 It is preferably at least one mass value selected from the group consisting of 2.6209, 1016.6113, 1020.4817, 1206.5305, 1207.5571, 1465.6184, 1466.6096, 1467.5969, 2450.9701, 2451.9662 and 2452.9546 m / z (with an error range of ± 0.1 m / z).

The cancer patient cases other than the plurality of colorectal cancer among the second type discrimination cases include at least one cancer patient case, non-Hodgkin lymphoma (NHL) patient case, and gastric cancer patient case Wherein mass values of the second type of colon cancer diagnostic low mass ions are 60.0476, 138.0540, 172.6653, 173.1158, 179.1451, 191.1277, 279.0855, 280.0895, 280.2642, 281.1440, 296.2574, 312.3248, 332.3224, 333.3324, 369.3406, 465.3161, At least one mass value selected from the group consisting of 486.6356, 488.6882, 544.8908, 551.3287, 566.8737, 707.3475 and 733.3569 m / z (with an error range of +/- 0.1 m / z).

Wherein the discrimination score includes a first type discrimination score by the first kind low mass ions and a second type discrimination score by the second kind low mass ions and the first type discrimination score is larger than CS ₁₁ . If the second type discrimination score is greater than CS ₂₁ , the discrimination object is judged to be colon cancer positive. If the first type discrimination score is smaller than CS ₁₂ or the second type discrimination score is smaller than CS ₂₂ It is possible to judge the discrimination object as colon cancer negative. Preferably, the CS ₁₁ , CS ₁₂ , CS _21, and CS ₂₂ are zero.

The low-mass ionic group that distinguishes colon cancer patients and normal individuals obtained using the above-mentioned colon selection diagnostic ion selection means may include fibrinogen or fibrinogen alpha chain.

Transthyretin may be included in the low mass ion group which distinguishes between the colon cancer patient and the normal person.

Wherein the first discrimination score calculating means comprises: A normalization module for normalizing the peak intensities of the low mass ion mass spectra of the training candidate set; A scaling module for scaling the normalized peak intensities; And a discrimination score calculation module for calculating the discrimination score by performing the biostatistical analysis on the scaled peak intensities.

The scaling module comprising: It is preferable to perform Pareto scaling.

Wherein the discrimination score calculation module comprises: The above biostatistical analysis can be performed using principal component analysis-based linear discriminant analysis (PCA-DA).

Wherein the discrimination score calculation module comprises: The discriminant score can be calculated using both the factor load value obtained by the linear discriminant analysis based on the principal component analysis and the scaled peak intensity together.

The factor load value calculation means comprising: Performing a biostatistical analysis on the sorted mass spectrum and selecting training cases that satisfy training conditions among the colorectal cancer patients and non-patient cases based on the biostatistical analysis results as a first training set 1 training set selection means, and may calculate an argument load value from the first training set.

Wherein the first training set selection means comprises: Patients with colorectal cancer and non-patient cases when the sensitivity according to the biostatistical analysis is equal to or greater than the threshold CN ₁ and the specificity is equal to or greater than the threshold CN ₂ may be set as the first training set. The threshold values CN ₁ and CN ₂ are preferably 1s.

And the second discrimination score calculating means comprises: A normalization module for normalizing the peak intensities of the low mass ion mass spectrum of the object to be discriminated; A scaling module for scaling the normalized peak intensities; And a discrimination score calculation module for calculating the discrimination score through the scaled peak intensity and the factor load value.

The scaling module comprising: It is preferable to perform Pareto scaling.

Wherein the discrimination score calculation module comprises: The discrimination score can be calculated through the scaled peak intensity of the low mass ions for colorectal cancer diagnosis and the factor load value.

The colorectal cancer determining means comprises: The colorimetric information of the discrimination subject is judged to be positive if the discrimination score is greater than the reference value CS, and if the discrimination score is smaller than the reference value CS, The colon cancer information of the discrimination object can be determined. The reference value CS is preferably zero.

The colorectal cancer determining means comprises: The colorectal cancer information of the discrimination object can be judged from the average value of a plurality of the discrimination scores calculated for a plurality of the low mass ion mass spectra detected by repeatedly measuring the biological sample to be discriminated.

The colon cancer diagnostic ion selecting means comprises: Candidate ion set selection means for selecting candidate low-mass ions satisfying a candidate condition from the selected first training set as a candidate ion set; And final ion set selection means for selecting low mass ions for colorectal cancer diagnosis as a final ion set based on the discrimination performance of individual candidate or low mass ions of the selected candidate ion set or combination thereof.

The candidate ion set selection means comprises: Wherein a peak intensity of the low mass ion for each of the training cases and an absolute value of a product of the low mass ion charge accumulation value obtained through the biostatistical analysis is greater than a threshold CT ₁ And a first low mass ion selection module to select for each training case. The threshold CT ₁ is preferably 0.1.

The candidate ion set selection means comprises: And a candidate ion cluster pre-selection module for selecting the second low-mass ions among the first low-mass ions as the candidate ion cluster, the second low-mass ions being commonly present in cases of CT ₂ percent or more of the thresholds of the entire training cases . The threshold CT ₂ is preferably 50.

The candidate ion set selection means comprises: A sensitivity and specificity calculating module for calculating a discrimination score indicating positive or negative for colorectal cancer in each training case using the second low-mass ions, and calculating sensitivity and specificity according to the discrimination score; And changing at least one of CT ₁ and CT ₂ when the sensitivity is smaller than a threshold value CN ₃ or the specificity is smaller than a threshold value CN ₄ and repeating the process to select the candidate ion set Ion collection final selection module. The thresholds CN ₃ and CN ₄ are preferably 0.9.

The criterion of the discrimination performance of the final ion set selection means is: Wherein the sum of the sensitivity and the specificity among the candidate low mass ions is larger than the reference value or the sum of the sensitivity and the specificity of each combination of the candidate low mass ions is greater than the sum of the sensitivity and the specificity, 1 criteria.

The criterion of the discrimination performance of the final ion set selection means is: And a second criterion in which the number of the candidate low mass ions among the combinations of the candidate low mass ions is selected as the least combination of the comparison targets.

The criterion of the discrimination performance of the final ion set selection means is: The difference between the minimum discrimination score of the true positive case and the maximum discrimination score of the true negative case among the combinations of the candidate low mass ions is the third criterion for selecting the largest combination of the comparison object combinations Wherein the discrimination score is calculated through the scaled peak intensity of the candidate low mass ions and the factor load value and is preferably positive or negative for colorectal cancer.

Said final ion set selection means comprising: Wherein the candidate low-mass ions included in the candidate ion set include high-sensitivity low-mass ions whose sensitivity is higher than the specificity, and high sensitivity that aligns the high-sensitivity low-mass ions in descending order of the sensitivity and the specificity The set {Sns ₁ , Sns ₂ , Sns ₃ ... Sns _I }, a high specificity set {Spc ₁ , Sns _I } comprising the high specificity low mass ions whose specificity is higher than the sensitivity, and sorting the high specific low mass ions in descending order of the sensitivity and the specificity, Spc ₂ , Spc ₃ ... Spc _J }; The high-sensitivity low-mass ions {Sns ₁ , Sns ₂ , Sns ₃ ... Sns _L } and the high-order L low-mass ions {Spc ₁ , Spc ₂ , Spc ₃ ... Of the candidate combination consisting of Spc _L} of two or more of the low mass ions the first reference and the second reference and the third the combination selected by the criterion for the discrimination performance by at least one of the reference biomarker A biomarker group pre-selection module to be selected as a group; Wherein said biomarker group comprises at least one of the above-mentioned high-sensitivity low-mass ions at the next higher-order M in said hypertonic sensitivity set and at least one of said high-specific low- A re-selection of a biomarker group reassigning the combination selected by the criterion of the discrimination performance by at least one of the first criterion, the second criterion and the third criterion among the candidate combinations added to the group to the biomarker group module; And a biomarker group final selection module for finalizing the biomarker group by repeating the re-selection process until there is no next-low mass ion in the hyper-sensitivity set and the high-specificity set.

Said final ion set selection means comprising: The remaining candidate ion sets excluding the combination of the low-mass ions selected as the biomarker group obtained in the biomarker group final selection module in the candidate ion set are repeatedly subjected to the selection process of the three biomarker groups, A biomarker group addition selecting module for further selecting the additional biomarker group until the L or less mass ions remain in the hyperhidimetic sensitivity set or the high specificity set; And selecting low mass ions of the combination of the upper CK biomarker groups based on the accuracy of the true positive and negative negative determinations among the biomarker group and the additional biomarker groups as the low mass ion for colon cancer diagnosis And a low mass ion final selection module for cancer diagnosis. Preferably, the L value is 2, the M value is 1, and the CK value is 1, 2, or 3.

The final ion set selection means preferably performs a low mass ion selection process on a training set in which a second training set independent of the first training set is added to the first training set.

The diagnosis apparatus for colorectal cancer according to the present invention has an advantage that the analysis cost is very low in the case of CRC diagnosis, the analysis time is short, and large-scale analysis is possible. In brief, the mass spectra of low-mass ions in the blood are measured, peak intensities corresponding to the mass values of low mass ions for CRC diagnosis are extracted, and positive / negative information for CRC is obtained by simple calculation can do.

In addition, it was confirmed that sensitivity, specificity, positive predictive value, and negative predictive value were more than 85% in the discrimination of not only the training set but also the verification set when the discrimination performance was excellent and robust. In addition, it can be usefully applied to various diseases by changing CRC patient and non-patient aggregation to other disease patient and non-patient aggregation.

In comparison with FOBT, which uses feces as an analytical sample, CRC can be used as an analytical sample. Therefore, analysis can be performed in addition to other tests, so that CRC information can be conveniently and quickly provided . In the case of using low mass ions for CRC diagnosis, compared with the discrimination performance of the conventional FOBT, it showed comparable performance in terms of specificity and superior performance in terms of sensitivity.

1 and 2 are views for explaining a conventional technique,
1 is a flow chart illustrating a process for determining a CRC using a low mass ion mass spectrum in accordance with the prior art,
FIG. 2 is a graph showing a result of determination of a set consisting of 133 CRC patients and 153 normal controls according to the prior art.
3 to 20 are views for explaining a colon cancer diagnosis apparatus according to a preferred embodiment of the present invention,
FIG. 3 is a block diagram showing a colon cancer diagnostic apparatus according to a preferred embodiment of the present invention. FIG.
FIG. 4 is a block diagram showing the colon cancer diagnosis unit of FIG. 3 in more detail;
FIG. 5 is a block diagram showing the first discrimination score calculating means of FIG. 3 in more detail;
FIG. 6 is a block diagram showing the second discrimination score calculating means of FIG. 3 in more detail;
FIG. 7 is a block diagram showing the colon cancer diagnostic ion selecting means of FIG. 3 in more detail;
FIG. 8 is a block diagram showing the candidate ion set selection means of FIG. 7 in more detail;
Fig. 9 is a block diagram showing the final ion set selection means of Fig. 7 in more detail;
FIG. 10 is a flowchart for explaining a process of selecting a first training set A ₀ having sensitivity and specificity of a predetermined value according to the present invention and calculating a weight-by-
11 is a flow chart for explaining a process of applying a discrimination equation to a sample to be discriminated,
FIG. 12 is a graph showing the result of determination of the set A ₁ by the weight-by-mass weight calculated in the first training set A ₀₁ ,
Figure 13 A graph showing the result of determination of the set A ₁ by the weight-by-mass weight calculated in the first training set A ₀₂ ,
FIG. 14 is a flowchart for explaining the process of constructing the preliminary discrimination formula according to the present invention,
FIG. 15 is a graph showing the results of determination of the set A ₁ in the first type discrimination formula,
16 is a cross- A graph showing the result of judging the set A ₁ by the second type preliminary discrimination formula,
17 is a flowchart for explaining the process of configuring the final discrimination formula according to the present invention,
FIG. 18 is a graph showing the results of determining the mean scores (mean DS) after calculating the discrimination scores according to the final discrimination formulas of the mass spectra measured five times repeatedly in the set A,
FIG. 19 is a graph showing a result of determination of the set B repeatedly measured five times by calculating the average discrimination score according to the final discrimination formula of the present invention,
20A is a graph showing the results of identifying 1465.6184 m / z among the mass values of the low-mass ions for the first kind of CRC diagnosed by the method according to the present invention,
FIG. 20B is a graph showing the results of identifying 2450.9701 m / z among the mass values of the low-mass ions for the first kind of CRC diagnosed by the method according to the present invention.

1. Definition of Terms

In the present invention, the "biological sample" includes samples such as whole blood, serum, plasma, urine, feces, sputum, saliva, tissues, cells, cell extracts, , But is not limited thereto. In the examples described below, sera from patients or non-patients was used as a biological sample.

In the present invention, "peak intensity" refers to a value obtained from a MALDI-TOF mass spectrometer and has a correlation with an amount of a mass ion corresponding to a peak.

In the present invention, "normalization" refers to matching the range of data or making the distribution similar, and may be normalized using mean, median, Various known methods may be applied in some cases. In this embodiment, the subtotal of the peak intensities for each sample is obtained, and the average of the sub-fields for each sample is obtained. Then, each peak intensity is normalized by multiplying the peak intensity by the sample magnification factor so that the subtotal of the peak intensities for each sample coincides with this average . That is, after the normalization as described above, the subtotal of the peak intensities of the respective samples becomes equal.

In the present invention, "Pareto scaling" refers to subtracting the mean value of each mass ion at each peak intensity normalized and dividing by the square root of the standard deviation. In autoscaling, a more general scaling method, data size information is completely canceled by dividing by standard deviation. In Pareto scaling, partial amplification of data size information is avoided to prevent amplification of noise. There is an advantage that it can be.

In the present invention, "weight" refers to a factor that adjusts the numerical magnitude of the data after multiplying the weight by proportional to its significance from a statistical point of view, which is obtained as a result of principal component analysis based linear discriminant analysis An example of weighting factor loadings for mass ions is an example.

In the present invention, "low mass ion" means an ion having a mass value obtained using a MALDI-TOF mass spectrometer or the like, which is smaller than 1500 m / z. Some of the low mass ions for CRC diagnosis have higher mass values than this range, but most of them fall within this range, so we will use the name of low mass ion as it is. That is, the limit of 1500 m / z is used as a rough value rather than a definite value.

In the present invention, the mass value measured by a MALDI-TOF mass spectrometer includes an error range of "± 0.05 m / z. This is because some error may occur in the mass value measurement depending on the experimental environment. For example, it should be understood that the mass value of 1467.5969 m / z described in the claims actually has a range of 1467.5469 m / z to 1467.6469 m / z. Depending on the experimental environment, the error range can be "± 0.1 m / z".

Note that in the present invention, the mass value measured by a MALDI-TOF mass spectrometer is a mass value obtained in a positive mode of a MALDI-TOF mass spectrometer.

In the present invention, the sign of the weight vector is adjusted to be positive when the discrimination score is a positive number and negative when it is negative. The factor load vector in the principal component analysis mathematically corresponds to an eigenvector, and the sign of this vector can be arbitrarily set. In other words, even if we change the code by multiplying the computed mass-coefficient-by-factor loadings by -1 as a whole, it is mathematically equivalent to the same eigenvalue problem, but if the discriminant score is negative, And the case of positive is judged by voice. In the present invention, the sign of the eigenvector is adjusted to be positive when the discrimination score is positive and negative when it is negative, but it is noted that the scope of the present invention is not limited thereto.

In the present invention, the &quot; discrimination score &quot; is a value calculated from a biostatistical analysis of a mass spectrum extracted from a biological sample, and based on this, the positive or negative of a specific cancer can be determined. The determination method may be a simple method of determining whether the calculated determination score is greater than a specific reference value or a function of inputting the calculated determination score and outputting the determination result as an output.

In the embodiment of the present invention, the term "discrimination score" is specified and used. However, the discrimination score defined in the present invention may be expressed in various types of terms such as a discrimination level and a discrimination value. Therefore, the term &quot; discriminant score &quot; defined in the present invention is not limited to the term &quot; discriminant score &quot; in the dictionary meaning, but may be the sum of the discriminant level, discriminant value, Should be construed as including.

In the present invention, the term &quot; discriminant performance &quot; means basically indicating numerical values such as sensitivity, specificity, positive predictive value, negative predictive value, and accuracy. It also means a value calculated as a function of these indicators. For example, each value of sensitivity, specificity, positive predictive value, negative predictive value, and accuracy can be used as the discriminant performance, or the sum of sensitivity and specificity, sum of sensitivity and positive predictive value, The sum of two or more such as the sum of the accuracy may be used as the discrimination performance.

Hereinafter, the present invention will be described in more detail by way of examples. It is to be understood that the embodiments described are intended to illustrate the present invention and the scope of the present invention is not limited to the embodiments.

2. Example

3 to 20 are views for explaining a colon cancer diagnosis apparatus according to a preferred embodiment of the present invention.

FIG. 3 is a block diagram illustrating a colon cancer diagnostic apparatus according to an exemplary embodiment of the present invention. As shown in FIG. 3, the diagnostic apparatus for colon cancer according to the present invention includes a plurality of colorectal cancer patients and non- Mass ion detector 1000 for detecting the mass spectrum of low-mass ions from the mass spectrometer 1000; A colon cancer diagnosis unit 2000 for comparing and analyzing the mass spectral patterns of the low-mass ions to determine colon cancer diagnosis information; And a display unit 3000 for converting the colorectal cancer diagnosis information determined by the colorectal cancer diagnosis unit 2000 into an outputable form and displaying the colorimetric cancer diagnostic information.

The low mass ion detector 1000 includes: As a specific example of information of low mass ions from the biological sample, peak intensity can be detected to extract the mass spectrum of the low mass ion. For this purpose, the low mass ion detector 1000 may include a mass analyzer.

The display unit 3000 includes: The detected cancer diagnosis information can be converted into various output forms such as letters, numerals, and graphics in a display device such as a monitor screen, a liquid crystal screen of a portable terminal, and the like.

FIG. 4 is a detailed block diagram of the colon cancer diagnosis unit of FIG. 3, wherein the colon cancer diagnosis unit 2000 comprises: First alignment means (2100) for aligning low mass ion mass spectra of said colorectal cancer patients and non-patient cases which are training candidate sets; A first discrimination score calculation means 2200 for calculating a discrimination score by performing a biostatistical analysis on the sorted mass spectrum; A factor load calculation means 2300 for calculating sensitivity and specificity according to the discrimination score and calculating a low mass ion factor load value by selecting a first training set based on the sensitivity and specificity; An ion selecting means (2400) for diagnosing colorectal cancer which selects low mass ions for diagnosing colorectal cancer based on the discrimination performance among candidate low mass ions satisfying the candidate condition; Second alignment means (2500) for aligning the low mass ion mass spectrum of the identified biological sample with the first training set; A second discrimination score calculation means (2600) for calculating a discrimination score from the peak intensity of the low mass ion of the discrimination object and the factor load value; And colorectal cancer determination means (2700) for determining the discriminant as positive or negative for colorectal cancer based on the discrimination score.

The colon cancer diagnostic ion selecting means 2400 comprises: The plurality of colorectal cancer patients and non-patient cases are classified into first type discrimination cases each comprising a plurality of colon cancer patient cases and a plurality of normal case cases, And the second kind discrimination cases are respectively executed for the first type discrimination cases and the second type discrimination cases, and the low-mass ions for diagnosing colorectal cancer are divided into the first type discrimination cases and the second kind discrimination cases, The low-mass ions for diagnosis of the first kind of colorectal cancer and the low-mass ions for diagnosis of the second kind of colorectal cancer in the second classification case are distinguished.

The parameter value calculation means (2300) comprises: Performing a biostatistical analysis on the sorted mass spectrum and selecting training cases that satisfy training conditions among the colorectal cancer patients and non-patient cases based on the biostatistical analysis results as a first training set 1 training set selection means 2310 and may calculate an argument load value from the first training set.

The first training set selection means (2310) comprises: The case of colon cancer patients and non-patient cases when the sensitivity according to the biostatistical analysis is equal to or greater than the threshold CN ₁ and the specificity is equal to or greater than the threshold CN ₂ may be set as the first training set. Here, the threshold values CN ₁ and CN ₂ are preferably 1s.

The colorectal cancer determining means (2700) comprises: The colorimetric information of the discrimination subject is judged to be positive if the discrimination score is greater than the reference value CS, and if the discrimination score is smaller than the reference value CS, The colon cancer information of the discrimination object can be determined. Here, the reference value CS is preferably zero.

The colorectal cancer determining means (2700) comprises: The colorectal cancer information of the discrimination object can be judged from the average value of a plurality of the discrimination scores calculated for a plurality of the low mass ion mass spectra detected by repeatedly measuring the biological sample to be discriminated.

Fig. 5 is a block diagram showing the first discrimination score calculating means of Fig. 4 in more detail. As shown in Fig. 5, the first discrimination score calculating means 2200 includes: A normalization module (2210) for normalizing the peak intensities of the low mass ion mass spectra of the training candidate set; A scaling module (2220) for scaling the normalized peak intensities; And a discrimination score calculation module 2230 for calculating the discrimination score by performing the biostatistical analysis on the scaled peak intensities.

The scaling module 2220 includes: Pareto scaling can be performed. The discrimination score calculation module 2230 includes: The biostatistical analysis can be performed using linear discriminant analysis based on principal component analysis. The discrimination score calculation module 2230 includes: The discriminant score can be calculated using both the factor load value obtained by the linear discriminant analysis based on the principal component analysis and the scaled peak intensity together.

FIG. 6 is a block diagram showing the second discrimination score calculating means of FIG. 4 in more detail. As shown in FIG. 6, the second discrimination score calculating means 2600 includes: A normalization module (2610) for normalizing the peak intensities of the low mass ion mass spectrum of the object to be discriminated; A scaling module (2620) for scaling the normalized peak intensities; And a discrimination score calculation module 2630 for calculating the discrimination score through the scaled peak intensity and the factor load value.

Here, the scaling module 2620 may perform pareto scaling. Also, the discrimination score calculation module 2630 includes: The discrimination score can be calculated through the scaled peak intensity of the low mass ions for colorectal cancer diagnosis and the factor load value.

FIG. 7 is a detailed block diagram of the colon cancer diagnostic ion selecting means of FIG. 4. As shown in FIG. 7, the colon cancer diagnostic ion selecting means 2400 includes: Candidate ion set selection means (2410) for selecting candidate low-mass ions satisfying a candidate condition from the selected first training set as a candidate ion set; And a final ion set selection means (2420) for selecting low mass ions for colorectal cancer diagnosis as a final ion set based on the discrimination performance of the candidate low mass ions of the selected candidate ion set, individually or in combination .

The criterion of the discrimination performance of the final ion set selection means 2420 is: Wherein the sum of the sensitivity and the specificity among the candidate low mass ions is larger than the reference value or the sum of the sensitivity and the specificity of each combination of the candidate low mass ions is greater than the sum of the sensitivity and the specificity, 1 criteria.

The criterion of the discrimination performance of the final ion set selection means 2420 is: And a second criterion in which the number of the candidate low mass ions among the combinations of the candidate low mass ions is selected as the least combination of the comparison targets.

The criterion of the discrimination performance of the final ion set selection means 2420 is: The difference between the minimum discrimination score of the true positive case and the maximum discrimination score of the true negative case among combinations of the candidate low mass ions is the third criterion for selecting the largest combination of the comparison target combinations Wherein the discrimination score is calculated through the scaled peak intensity of the candidate low mass ions and the factor load value and can be positive or negative for colorectal cancer.

The final ion set selection means 2420 may perform a low mass ion selection process on a training set in which a second training set independent of the first training set is added to the first training set.

FIG. 8 is a block diagram showing in more detail the candidate ion set selection means of FIG. 7, wherein the candidate ion set selection means 2410 includes: Wherein a peak intensity of the low mass ion for each of the training cases and an absolute value of a product of the low mass ion charge accumulation value obtained through the biostatistical analysis is greater than a threshold CT ₁ And a first low mass ion selection module 2411 for selecting the training case for each training case. Here, the threshold CT ₁ is preferably 0.1.

The candidate ion set selection means (2410) comprises: And a candidate ion set pre-selection module (2412) for selecting the second low-mass ions as the candidate ion set common to the cases of not less than ₂ percent of the threshold CT of the entire training cases among the first low-mass ions can do. Here, the threshold value CT ₂ is preferably 50.

The candidate ion set selection means (2410) comprises: A sensitivity and specificity calculation module 2413 for calculating a discrimination score indicating positive or negative for colorectal cancer in each training case using the second low-mass ions, and calculating sensitivity and specificity according to the discrimination score; And changing at least one of CT ₁ and CT ₂ when the sensitivity is smaller than a threshold CN ₃ or the specificity is smaller than a threshold CN ₄ and repeating the process to select the candidate ion set And an ion collection final selection module 2414. Here, the threshold values CN ₃ and CN ₄ are preferably 0.9.

FIG. 9 is a block diagram showing in more detail the final ion set selection means of FIG. 7, wherein the final ion set selection means 2420 includes: Wherein the candidate low-mass ions included in the candidate ion set include high-sensitivity low-mass ions whose sensitivity is higher than the specificity, and wherein the low-sensitivity low-mass ions are arranged in descending order of the sensitivity and the sum of the specificities Sensitivity set {Sns ₁ , Sns ₂ , Sns ₃ ... Sns _I }, a high specificity set {Spc ₁ , Sns _I } comprising the high specificity low mass ions whose specificity is higher than the sensitivity, and sorting the high specific low mass ions in descending order of the sensitivity and the specificity, Spc ₂ , Spc ₃ ... An ion sorting module 2421 for dividing the ion species into Spc _J ; The high-sensitivity low-mass ions {Sns ₁ , Sns ₂ , Sns ₃ ... Sns _L } and the high-order L low-mass ions {Spc ₁ , Spc ₂ , Spc ₃ ... Of the candidate combination consisting of Spc _L} of two or more of the low mass ions the first reference and the second reference and the third the combination selected by the criterion for the discrimination performance by at least one of the reference biomarker A biomarker group preliminary selection module 2422 for selecting as a group; Wherein said biomarker group comprises at least one of the above-mentioned high-sensitivity low-mass ions at the next higher-order M in said hypertonic sensitivity set and at least one of said high-specific low- A re-selection of a biomarker group reassigning the combination selected by the criterion of the discrimination performance by at least one of the first criterion, the second criterion and the third criterion among the candidate combinations added to the group to the biomarker group Module 2423; And a biomarker group final selection module (2424) for final selection of the biomarker group by repeating the re-selection process until there is no next-low mass ion in the high sensitivity set and the high specificity aggregate have.

The final ion set selection means (2420) comprises: The above-mentioned three biomarker group selection process is repeated for the remaining candidate ion sets excluding the combination of the low-mass ions selected as the biomarker group obtained in the biomarker group final selection module 2424 in the candidate ion set A biomarker group addition selecting module (2425) for selecting the biomarker group, further selecting the additional biomarker group until the L or less mass ions remain in the high sensitivity set or the high specificity aggregate; And selecting low mass ions of the combination of the upper CK biomarker groups based on the accuracy of the true or negative determination of the biomarker group and the additional biomarker groups as the low mass ion for diagnosing the colorectal cancer And a low mass ion final selection module 2426 for cancer diagnosis. Here, the L value is 2, the M value is 1, and the CK value is preferably 1, 2, or 3.

The colorectal cancer diagnosis apparatus of the present invention may be constituted by hardware or software by a program structure as shown in FIG. 4. Hereinafter, referring to the flowchart of the attached flowchart, for example, The colorectal cancer diagnosing apparatus will be described in detail.

2-1. Sample Preparation - Serum Collection Target

133 CRC patients in Table 1, 153 normal controls in Table 2, 111 BRC patients in Table 5, 36 non-Hodgkin lymphoma (NHL) patients in Table 6, and 29 patients in Table 7 Serum was collected from a group of GC patients.

BRC Sex Age
year Node ER ER% PR PR% HER2 Tumor Size cm BRC-A1 F 48 - 5 33-66% 6 33-66% 2 - BRC-A2 F 35 - 6 33-66% 6 33-66% One - BRC-A3 F 45 pN1a 5 33-66% 5 33-66% 0 1.5 BRC-A4 F 61 - 0 0% 0 0% 2 - BRC-A5 F 70 pN0 (sn) 0 0% 0 0% One &Lt; 0.1 BRC-A6 F 58 ypN0 3 <10% 3 10-33% 3 0.5 BRC-A7 F 49 ypN0 (i +) 0 0% 0 0% 2 1.9 BRC-A8 F 49 ypN2a 0 0% 0 0% One 2.5 BRC-A9 F 39 pN1a 6 33-66% 7 > 66% One 2.2 BRC-A10 F 48 ypN2a 6 33-66% 4 <10% 3 5.8 BRC-A11 F 39 - 0 0% 0 0% One - BRC-A12 F 56 pN1a 6 33-66% 6 33-66% 0 2.8 BRC-A13 F 59 pN0 (sn) 6 33-66% 2 <10% One 2.3 BRC-A14 F 31 pN1a 5 33-66% 4 10-33% One 2.2 BRC-A15 F 46 pN3a 6 33-66% 6 33-66% One 3.5 BRC-A16 F 56 - 7 > 66% 4 10-33% One - BRC-A17 F 55 - 0 0% 0 0% 2 - BRC-A18 F 46 pN0 0 0% 0 0% 0 1.5 BRC-A19 F 60 ypN0 0 0% 0 0% 3 1.9 BRC-A20 F 49 pN0 (sn) 5 33-66% 2 <10% 2 1.5 BRC-A21 F 55 pN1m 0 0% 0 0% 3 1.8 BRC-A22 F 65 pN0 6 33-66% 6 33-66% 0 1.7 BRC-A23 F 35 ypN2a 6 66% 4 10-33% 2 2.6 BRC-A24 F 46 pN1a 6 33-66% 6 33-66% 3 2.5 BRC-A25 F 45 pN0 (sn) 6 33-66% 6 33-66% One 0.8 BRC-A26 F 42 pN0 (sn) 3 10-33% 6 33-66% 0 One BRC-A27 F 58 pN0 (sn) 6 33-66% 6 33-66% One 1.5 BRC-A28 F 62 pN1a 0 0% 0 0% 2 2.2 BRC-A29 F 61 - 0 0% 0 0% One - BRC-A30 F 60 - - - - - - - BRC-A31 F 51 - - - - - - - BRC-A32 F 42 pN0 7 > 66% 7 > 66% 2 - BRC-A33 F 43 pN0 (sn) 3 10-33% 4 10-33% 0 2.3 BRC-A34 F 60 pN0 (sn) 0 0% 0 0% One 2.3 BRC-A35 F 61 - 6 33-66% 0 0% 2 - BRC-A36 F 61 pN0 (sn) 0 0% 2 <10% 2 1.8 BRC-A37 F 49 - - - - - - - BRC-A38 F 44 pN0 6 33-66% 7 > 66% One 1.2 BRC-A39 F 72 pN0 (sn) 0 0% 0 0% 0 1.8 BRC-A40 F 48 pN0 (sn) 5 33-66% 4 10-33% One 0.8 BRC-A41 F 44 pN0 5 33-66% 7 > 66% One 2 BRC-A42 F 41 pN2a 5 33-66% 6 33-66% One 4 BRC-A43 F 58 pN0 6 33-66% 0 0% 2 &Lt; 0.1 BRC-A44 F 42 - 5 33-66% 6 33-66% 2 - BRC-A45 F 44 pN1a 4 10-33% 2 <10% 2 5.5 BRC-A46 F 62 pN0 (sn) 7 > 66% 0 0% 0 2 BRC-A47 F 47 pN0 6 33-66% 6 33-66% 2 2.4 BRC-A48 F 52 pN1a 6 33-66% 0 0% 3 1.8 BRC-A49 F 44 pN0 (sn) 6 33-66% 0 0% 0 2 BRC-A50 F 49 pN0 (sn) 2 <10% 2 <10% 3 0.4 BRC-A51 F 46 pN0 (sn) 6 33-66% 5 33-66% One 0.7 BRC-A52 F 58 pN0 (sn) 7 > 66% 5 33-66% One 2.3 BRC-A53 F 64 pN1a 6 33-66% 7 > 66% One 2 BRC-A54 F 47 - 6 33-66% 6 33-66% 2 - BRC-A55 F 74 pN1a 6 33-66% 6 33-66% One 1.8 BRC-A56 F 64 pN0 (sn) 0 0% 0 0% One 2.2 BRC-A57 F 40 ypN1a 6 33-66% 6 33-66% One 3.5 BRC-A58 F 43 pN0 6 33-66% 6 33-66% 2 2.5 BRC-A59 F 43 ypN0 0 0% 0 0% 2 - BRC-A60 F 42 pN0 0 0% 0 0% 0 2.3 BRC-A61 F 37 pN0 (i +) 6 33-66% 6 33-66% One One BRC-A62 F 50 pN1a 6 33-66% 6 33-66% One One BRC-A63 F 57 pN0 (sn) 6 33-66% 96 33-66% One 1.4 BRC-A64 F 38 ypN0 0 0% 0 0% One 2 BRC-A65 F 67 - 6 33-66% 2 <10% One - BRC-A66 F 42 pN0 (sn) 6 33-66% 6 33-66% 2 0.5 BRC-A67 F 46 pN0 (sn) 6 33-66% 6 33-66% One One BRC-A68 F 48 pN2a 4 10-33% 4 10-33% 3 2.5 BRC-A69 F 58 pN0 2 <10% 0 0 One 0.5 BRC-A70 F 53 pN0 (sn) 0 0% 0 0% 3 &Lt; 0.1 BRC-A71 F 56 - 0 0% 0 0% 0 - BRC-A72 F 45 pN0 (sn) 6 33-66% 6 33-66% 2 &Lt; 0.1 BRC-A73 F 59 pN0 (sn) 5 33-66% 0 0% 2 1.4 BRC-A74 F 40 ypN1a 2 <10% 0 0% 0 0.3 BRC-A75 F 34 pN0 (sn) 2 <10% 0 0% 2 2 BRC-A76 F 69 - 6 33-66% 6 33-66% One - BRC-A77 F 52 - - - - - - - BRC-A78 F 67 - - - - - - - BRC-A79 F 61 - 6 33-66% 2 <10% 0 - BRC-A80 F 38 pN1a 6 33-66% 5 33-66% One - BRC-A81 F 60 pN0 6 33-66% 3 10-33% One One BRC-A82 F 55 pN2a 5 33-66% 0 0% 2 2.2 BRC-A83 F 46 ypN0 5 33-66% 2 <10% One 1.5 BRC-A84 F 67 pN0 6 33-66% 6 33-66% One 2.8 BRC-A85 F 46 pN1a 6 33-66% 6 33-66% 2 0.7 BRC-A86 F 39 pN1m 6 33-66% 6 33-66% 2 2.5 BRC-A87 F 50 pN0 (sn) 4 10-33% 5 33-66% 0 One BRC-A88 F 31 pNlm (sec) 6 33-66% 6 33-66% One One BRC-A89 F 46 pN0 6 33-66% 7 > 66% One 1.2 BRC-A90 F 44 pN0 (sn) 6 33-66% 7 > 66% One 2.5 BRC-A91 F 40 pN0 0 0% 0 0% 0 - BRC-A92 F 40 - 6 33-66% 6 33-66% One - BRC-A93 F 56 - 7 > 66% 0 0 0 0.6 BRC-A94 F 48 pN1a 0 0% 0 0% 0 3 BRC-A95 F 39 pN0 (sn) 6 33-66% 6 33-66% One 3.5 BRC-A96 F 40 ypN1a 6 33-66% 4 10-33% 2 3 BRC-A97 F 48 pN0 (sn) 6 33-66% 6 33-66% 0 2.5 BRC-A98 F 59 - 7 > 66% 2 <10% One - BRC-A99 F 46 - 0 0% 0 0% 2 - BRC-A100 F 37 pN3a 6 33-66% 6 33-66% 2 0.6 BRC-A101 F 38 pN0 (sn) 6 33-66% 6 33-66% 2 0.3 BRC-A102 F 66 pN1a 6 33-66% 6 33-66% 0 1.5 BRC-A103 F 58 pN0 (sn) 0 0% 0 0% 2 1.7 BRC-A104 F 42 pN3a 5 33-66% 6 33-66% 0 1.8 BRC-A105 F 52 pN0 6 33-66% 6 33-66% 0 0.7 BRC-A106 F 46 pN0 (sn) 0 0% 2 <10% One 1.5 BRC-A107 F 42 pN0 (sn) 4 10-33% 6 33-66% One 0.6 BRC-A108 F 48 - - - - - - - BRC-A109 F 47 pN0 6 33-66% 2 <10% 2 3 BRC-A110 F 59 pN1a 6 33-66% 4 10-33% One 1.8 BRC-A111 F 56 - 0 0% 0 0% 3 -

NHL Sex Age
year Stage Involved Site Subtype IPI NHL-A1 M 65 3 multiple DLBL 3 NHL-A2 M 63 2 stomach DLBL One NHL-A3 M 65 3 multiple DLBL 3 NHL-A4 M 65 3 multiple Follicular L 2 NHL-A5 M 64 2 stomach DLBL 2 NHL-A6 M 52 3 multiple DLBL 2 NHL-A7 M 52 2 spleen, pancreatic LN DLBL One NHL-A8 M 52 3 multiple DLBL 2 NHL-A9 M 42 2 multiple DLBL 2 NHL-A10 M 44 One stomach DLBL One NHL-A11 M 44 2 cervical LN, tonsil DLBL 0 NHL-A12 M 40 4 multiple DLBL 3 NHL-A13 M 39 One nasal cavity NK / T cell L One NHL-A14 M 41 One inguinal LN ALCL 0 NHL-A15 F 57 4 multiple DLBL 3 NHL-A16 F 38 2 tonsil, neck LN DLBL 0 NHL-A17 F 56 One breast DLBL 0 NHL-A18 M 71 4 multiple Mantle cell L 3 NHL-A19 M 70 2 neck area LN DLBL One NHL-A20 M 80 2 stomach PTCL One NHL-A21 F 39 2 tonsil, neck LN DLBL 0 NHL-A22 F 38 4 multiple DLBL 3 NHL-A23 F 38 One stomach DLBL 0 NHL-A24 M 67 4 multiple DLBL 2 NHL-A25 M 67 3 multiple Burkitt's L 3 NHL-A26 F 73 One nasal cavity DLBL 2 NHL-A27 F 73 3 multiple DLBL 2 NHL-A28 F 41 One 0 DLBL 0 NHL-A29 M 49 3 multiple DLBL 3 NHL-A30 M 31 2 neck DLBL 0 NHL-A31 M 46 2 nasopharynx, tonsil DLBL 0 NHL-A32 M 71 - stomach r / o Lymphoma - NHL-A33 M 73 One nasal cavity DLBL One NHL-A34 M 73 4 tibia, leg (skin) DLBL 3 NHL-A35 M 72 2 stomach DLBL One NHL-A36 M 79 One nasal cavity malignant L 2

GC Sex Age
year CEA
ng / mL Stage GC Sex Age
year CEA
ng / mL Stage GC-A1 F 52 3.36 IV GC-A16 F 51 0.51 IV GC-A2 M 65 0.99 IV GC-A17 M 63 27.18 IV GC-A3 M 41 - ^a IV GC-A18 M 51 1.93 IV GC-A4 M 78 4.93 IV GC-A19 M 64 2.41 IV GC-A5 M 79 1.11 IV GC-A20 M 62 2.72 IV GC-A6 M 76 2.37 IV GC-A21 M 71 8.46 IV GC-A7 M 54 117.13 IV GC-A22 M 46 2.67 IV GC-A8 M 58 2.24 IV GC-A23 M 68 24.93 IV GC-A9 M 67 > 1500 IV GC-A24 M 68 3.23 IV GC-A10 F 71 1.92 IV GC-A25 M 57 41.32 IV GC-A11 F 42 <0.4 IV GC-A26 M 71 2.8 IV GC-A12 M 49 104.73 IV GC-A27 F 43 1.62 IV GC-A13 M 65 1.69 IV GC-A28 M 58 6.6 IV GC-A14 F 57 6.98 IV GC-A29 M 73 - IV GC-A15 M 55 2.03 IV

The first training set is composed of 462 persons as a set A ₁ , as described below, and its subset A ₀ is constructed. Through the biostatistical analysis of the first training set, the mass ion weight ), And a preliminary discriminant was obtained. In addition to this set A ₁ , 144 CRC patients in Table 8, 50 normal controls in Table 9, 25 BRC patients in Table 10, 15 NHL patients in Table 11, and 57 GC patients in Table 12 were grouped into groups A ₂ (The second training set). That is, from the preliminary candidates of the low-mass ions constituting the preliminary discrimination formula, when the low-mass ions for CRC diagnosis are identified according to the method described later, the union of the groups A ₁ and A ₂ , Respectively.

CRC Sex Age
year Stage Location Cell Type CEA
ng / mL CRC-A134 M 49 I Rectum AC 6.6 CRC-A135 F 60 III A-colon AC 30.4 CRC-A136 M 69 IV A-colon AC 33.5 CRC-A137 M 43 III A-colon MAC 77.6 CRC-A138 F 69 III Rectum AC 1.0 CRC-A139 M 72 III A-colon AC 2.4 CRC-A140 M 54 II Rectum AC 4.6 CRC-A141 M 58 II S-colon AC 2.9 CRC-A142 F 52 III S-colon AC 9.2 CRC-A143 M 52 III S-colon AC 3.2 CRC-A144 M 78 IV Rectum AC 4.1 CRC-A145 F 55 III Rectum AC 0.9 CRC-A146 M 65 II S-colon AC 1.7 CRC-A147 F 46 I S-colon AC 1.4 CRC-A148 M 77 III S-colon AC 2.5 CRC-A149 F 52 II S-colon AC <0.5 CRC-A150 F 47 III S-colon AC 1.5 CRC-A151 M 48 III S-colon AC 1.7 CRC-A152 F 76 II S-colon AC 2.2 CRC-A153 M 51 II S-colon ASC 8.6 CRC-A154 M 61 I Rectum AC 1.4 CRC-A155 M 56 II Rectum AC 3 CRC-A156 F 70 III S-colon MAC 36.0 CRC-A157 M 64 III A-colon AC 2.2 CRC-A158 F 54 III Rectum AC 5.5 CRC-A159 F 77 II Rectum AC 6.2 CRC-A160 M 53 III Rectum AC 1.4 CRC-A161 M 43 I Rectum AC 0.5 CRC-A162 M 81 III A-colon MAC 10.9 CRC-A163 F 52 III A-colon AC 1.2 CRC-A164 F 71 III A-colon AC 2.8 CRC-A165 M 84 III Rectum AC 15.0 CRC-A166 F 33 III D-colon AC 4.7 CRC-A167 F 68 III Rectum AC 3.3 CRC-A168 M 69 III Rectum AC 3.5 CRC-A169 F 61 III A-colon AC 2.8 CRC-A170 M 44 II T-colon AC 1.8 CRC-A171 F 82 II A-colon AC 2.8 CRC-A172 M 38 IV Rectum AC 2.1 CRC-A173 M 73 III Rectum AC 11.1 CRC-A174 M 64 III D-colon AC 8.2 CRC-A175 M 67 II A-colon AC 20.1 CRC-A176 M 72 II A-colon AC 3.4 CRC-A177 M 59 II S-colon AC 2.1 CRC-A178 M 53 I Rectum AC 3.5 CRC-A179 M 70 II Rectum AC 1.3 CRC-A180 M 55 II Rectum AC 22.0 CRC-A181 M 62 II Rectum AC 6.1 CRC-A182 M 64 III Rectum AC 4.8 CRC-A183 M 62 IV Rectum AC 25.3 CRC-A184 M 51 III Rectum AC 149.3 CRC-A185 F 45 II Rectum AC 2.7 CRC-A186 F 49 II Rectum AC 2.1 CRC-A187 F 45 0 Rectum AC 0.9 CRC-A188 M 62 III Rectum AC 2.4 CRC-A189 M 54 0 Rectum AC 6.9 CRC-A190 M 45 0 Rectum AC 7.4 CRC-A191 F 54 0 Rectum AC 3.6 CRC-A192 M 69 II Rectum AC 24.0 CRC-A193 M 51 I Rectum AC 2.7 CRC-A194 M 45 I Rectum AC 3.2 CRC-A195 M 67 I Rectum AC 2.9 CRC-A196 M 60 I Rectum AC 1.5 CRC-A197 M 49 0 Rectum AC 0.8 CRC-A198 M 71 I Rectum AC 9.8 CRC-A199 M 62 III Rectum AC 2.5 CRC-A200 M 54 II Rectum AC 4.6 CRC-A201 M 56 II Rectum AC 3.0 CRC-A202 F 71 III Rectum AC 6.7 CRC-A203 M 73 0 Rectum AC 61.5 CRC-A204 F 50 III Rectum AC 2.2 CRC-A205 F 49 0 Rectum AC 1.6 CRC-A206 F 42 III Rectum AC 9.9 CRC-A207 M 61 III Rectum AC 68.1 CRC-A208 F 72 II Rectum AC 8 CRC-A209 F 69 III Rectum AC 11.3 CRC-A210 M 58 II Rectum AC 5.3 CRC-A211 M 56 I Rectum AC 24.8 CRC-A212 M 72 III Rectum AC 1.4 CRC-A213 M 62 III Rectum AC 1.6 CRC-A214 M 55 II Rectum AC 2.4 CRC-A215 F 71 III Rectum AC 1.3 CRC-A216 M 59 III Rectum AC 2.8 CRC-A217 M 52 II Rectum AC 4.0 CRC-A218 M 47 III Rectum AC 2.3 CRC-A219 M 58 II Rectum AC 1.1 CRC-A220 M 60 0 Rectum AC 2.0 CRC-A221 M 64 I Rectum AC 2.0 CRC-A222 M 41 III Rectum AC 1.6 CRC-A223 M 48 I Rectum AC 0.8 CRC-A224 M 58 II Rectum AC 1.1 CRC-A225 M 61 I Rectum AC 2.6 CRC-A226 M 63 I Rectum AC 1.3 CRC-A227 F 52 II Rectum AC 1.6 CRC-A228 M 53 II Rectum AC 2.0 CRC-A229 M 64 I Rectum AC 2.0 CRC-A230 M 73 II Rectum AC 5.6 CRC-A231 M 41 III Rectum AC 1.6 CRC-A232 M 57 III Rectum AC 2.0 CRC-A233 M 48 I Rectum AC 0.8 CRC-A234 M 72 III Rectum AC 6.1 CRC-A235 F 67 0 Rectum AC 4.4 CRC-A236 F 66 II Rectum AC 4.8 CRC-A237 M 47 III S-colon AC 3.7 CRC-A238 M 40 III A-colon AC 1.2 CRC-A239 M 55 II D-colon AC 6.0 CRC-A240 F 73 I D-colon, T-colon AC 2.0 CRC-A241 F 69 I A-colon AC 5.0 CRC-A242 F 69 I A-colon AC 5.7 CRC-A243 F 74 II D-colon AC 12.5 CRC-A244 M 61 II S-colon MAC 1.9 CRC-A245 M 37 III Rectum AC 6.0 CRC-A246 M 60 III S-colon AC 5.4 CRC-A247 M 70 II S-colon AC 2.6 CRC-A248 M 68 III Rectum AC 13.2 CRC-A249 M 73 I Rectum AC 1.7 CRC-A250 M 82 III T-colon AC 2.1 CRC-A251 F 75 II Rectum AC 0.9 CRC-A252 F 57 I A-colon AC 1.5 CRC-A253 F 62 III S-colon AC 4.4 CRC-A254 M 73 II Rectum AC 15.5 CRC-A255 M 59 I S-colon AC 1.1 CRC-A256 F 74 III Rectum AC 31.0 CRC-A257 F 70 I A-colon AC 2.5 CRC-A258 M 74 II S-colon AC 15.4 CRC-A259 M 69 II Rectum AC 2.1 CRC-A260 M 61 II A-colon, T-colon AC 2.3 CRC-A261 M 73 I Rectum AC 1.9 CRC-A262 M 64 I Rectum AC 2.8 CRC-A263 M 69 II D-colon AC 5.0 CRC-A264 M 58 III Rectum AC 1.6 CRC-A265 M 73 II T-colon AC 2.6 CRC-A266 M 70 II A-colon AC 20.8 CRC-A267 M 56 IV Rectum AC 29.9 CRC-A268 F 70 II A-colon AC 5.9 CRC-A269 M 71 III S-colon AC 110.1 CRC-A270 M 47 III Rectum AC 13.7 CRC-A271 M 61 III Rectum AC 2.8 CRC-A272 F 77 II S-colon AC 1.5 CRC-A273 F 62 III Rectum AC 13.7 CRC-A274 M 61 II S-colon AC 2.3 CRC-A275 M 66 II S-colon AC 1.7 CRC-A276 M 64 III A-colon AC 1.0 CRC-A277 M 69 II S-colon AC 23.0

ASC: Adenosquamous carcinoma

Control Sex Age
year CEA
ng / mL Control Sex Age
year CEA
ng / mL CONT-A154 F 51 1.7 CONT-A179 F 63 1.8 CONT-A155 F 62 1.3 CONT-A180 F 65 1.1 CONT-A156 M 54 4.2 CONT-A181 M 64 1.4 CONT-A157 F 63 1.6 CONT-A182 F 55 4.8 CONT-A158 F 60 1.9 CONT-A183 M 63 2.6 CONT-A159 F 68 1.4 CONT-A184 F 52 4.1 CONT-A160 F 62 1.9 CONT-A185 M 51 4.0 CONT-A161 F 68 5.6 CONT-A186 M 59 2.0 CONT-A162 M 63 4.5 CONT-A187 M 68 4.6 CONT-A163 M 50 2.1 CONT-A188 M 50 5.0 CONT-A164 F 53 2.3 CONT-A189 F 64 <0.5 CONT-A165 M 60 3.3 CONT-A190 F 63 2.2 CONT-A166 M 64 1.8 CONT-A191 M 64 1.7 CONT-A167 M 63 3.4 CONT-A192 M 51 2.3 CONT-A168 F 63 1.1 CONT-A193 F 62 1.1 CONT-A169 M 53 2.0 CONT-A194 M 54 2.5 CONT-A170 F 51 2.0 CONT-A195 F 53 0.7 CONT-A171 M 57 3.3 CONT-A196 F 65 3.8 CONT-A172 M 61 2.8 CONT-A197 F 64 1.5 CONT-A173 F 68 1.4 CONT-A198 F 53 1.0 CONT-A174 F 52 1.5 CONT-A199 M 50 1.1 CONT-A175 M 60 4.6 CONT-A200 F 66 1.7 CONT-A176 M 55 2.2 CONT-A201 F 50 1.0 CONT-A177 M 55 1.8 CONT-A202 F 50 1.9 CONT-A178 M 56 2.2 CONT-A203 M 61 1.5

BRC Sex Age
year Node ER ER% PR PR% HER2 Tumor Size
cm BRC-A112 F 61 pN0 (i + 0) 7 >
95% 0 0% 0 4 BRC-A113 F 53 pN0 7 80% 5 25% 0 0.6 BRC-A114 F 49 pN0 3 20% 7 60% 0 0.3 BRC-A115 F 57 pN0 0 0% 0 0% 0 0.8 BRC-A116 F 68 pN0 0 0% 3 One% 3 1.2 BRC-A117 F 58 pN0 8 95% 4 40% 0 0.8 BRC-A118 F 40 - 8 95% 8 95% 0 - BRC-A119 F 29 pN0 8 95% 8 95% One 1.2 BRC-A120 F 40 - - - - - - - BRC-A121 F 61 pN0 (i + 0) 7 >
95% 0 0% 0 4 BRC-A122 F 40 - 8 95% 8 95% 0 - BRC-A123 F 43 pN0 0 0% 0 0% 3 0.7 BRC-A124 F 59 pN0 8 95% 8 95% 0 1.2 BRC-A125 F 45 PN2 7 95% 8 95% One 2.1 BRC-A126 F 55 pN0 0 0% 0 0% 3 1.8 BRC-A127 F 52 pN0 7 80-
90% 8 80-
90% 0 0.3 BRC-A128 F 59 pN0 8 95% 5 2 ~
3% One 1.3 BRC-A129 F 39 - 7 >
95% 7 70-
80% 0 - BRC-A130 F 39 pN0 0 0% 0 0% 3 1.1 BRC-A131 F 40 pN0 5 50-
60% 5 20-
30% 0 0.8 BRC-A132 F 46 pN0 7 95% 8 95% 0 4.9 BRC-A133 F 51 pN0 0 <1% 0 0% 0 0.9 BRC-A134 F 61 pN0 7 90% 8 90% 0 1.3 BRC-A135 F 48 pN0 0 0% 0 0% 0 0.6 BRC-A136 F 47 pN0 8 >
95% 8 95% 0 0.7

NHL Sex Age
year Stage Involved Site Subtype IPI NHL-A37 F 69 3 multiple ATCL 3 NHL-A38 F 72 One stomach DLBL 2 NHL-A39 M 24 4 multiple DLBL 3 NHL-A40 F 41 2 stomach DLBL One NHL-A41 F 48 4 multiple DLBL 3 NHL-A42 F 66 2 gum, submandibular DLBL One NHL-A43 F 61 2 stomach DLBL 3 NHL-A44 M 38 4 multiple Hodgkin L - NHL-A45 M 70 4 multiple DLBL 3 NHL-A46 M 37 One cervical LN DLBL 0 NHL-A47 M 64 4 multiple DLBL 4 NHL-A48 F 76 2 stomach DLBL One NHL-A49 M 34 2 neck, SCN DLBL One NHL-A50 M 25 4 multiple PTCL 2 NHL-A51 M 70 2 stomach DLBL One

GC Sex Age
year CEA
ng / mL Stage GC Sex Age
year CEA
ng / mL Stage GC-A30 M 70 1.26 III GC-A59 M 62 - II GC-A31 M 62 5.41 I GC-A60 M 67 - II GC-A32 M 58 - I GC-A61 F 64 - II GC-A33 M 62 1.34 I GC-A62 F 40 - II GC-A34 M 32 - I GC-A63 M 64 - II GC-A35 M 71 3.54 I GC-A64 M 68 - II GC-A36 M 56 2.83 I GC-A65 M 54 - II GC-A37 M 69 21.71 II GC-A66 F 52 - II GC-A38 F 62 - I GC-A67 M 59 - II GC-A39 M 52 1.86 I GC-A68 F 81 - II GC-A40 F 64 4.16 I GC-A69 M 46 - III GC-A41 M 59 - II GC-A70 M 62 - III GC-A42 M 61 10.41 IV GC-A71 M 51 - III GC-A43 F 68 5.56 III GC-A72 M 42 - III GC-A44 M 48 1.44 III GC-A73 M 81 - III GC-A45 F 80 - III GC-A74 F 81 - III GC-A46 M 66 - IV GC-A75 M 70 - III GC-A47 F 57 2.46 IV GC-A76 M 51 - III GC-A48 M 46 1.68 III GC-A77 M 68 - IV GC-A49 M 79 - I GC-A78 M 68 - IV GC-A50 M 81 - I GC-A79 F 33 - IV GC-A51 M 52 - I GC-A80 M 31 - IV GC-A52 M 53 - I GC-A81 M 52 - IV GC-A53 M 67 - I GC-A82 M 59 - IV GC-A54 M 61 - I GC-A83 M 56 - IV GC-A55 F 77 - I GC-A84 M 82 - IV GC-A56 F 74 - I GC-A85 F 52 - IV GC-A57 F 81 - I GC-A86 M 82 - IV GC-A58 F 55 - I

In addition, there were 143 CRC patients in Table 13, 13 healthy controls in Table 14, 25 BRC patients in Table 15, 15 NHL patients in Table 16, 55 GC patients in Table 17, 25 patients in Table 18 (OVC) and 19 patients with Tis or Advanced Adenoma (TA) in Table 19 were group B, and the test set was constructed. The OVC patients and the TA patients were included in the group of patients who were not reflected in the calculation of the mass ion specific weight or the low mass ion for CRC diagnosis.

CRC Sex Age
year Stage Location Cell Type CEA
ng / mL CRC-B1 F 51 II Rectum AC 6.7 CRC-B2 M 71 I Rectum AC 9.8 CRC-B3 F 47 I Rectum AC 3.9 CRC-B4 F 50 IV Rectum AC 62.0 CRC-B5 M 79 II S-colon AC 6.2 CRC-B6 F 54 I Rectum AC 1.6 CRC-B7 F 52 III S-colon AC 22.1 CRC-B8 M 61 III Rectum AC 128.1 CRC-B9 F 47 III S-colon AC 1.2 CRC-B10 M 73 I S-colon AC 7.1 CRC-B11 F 74 I S-colon AC 2.3 CRC-B12 M 71 III A-colon AC 8.2 CRC-B13 M 57 IV Rectum AC 6.4 CRC-B14 M 57 IV S-colon AC 41.7 CRC-B15 M 52 III S-colon AC 4.1 CRC-B16 F 64 III S-colon AC 6.8 CRC-B17 M 48 IV A-colon AC 59.4 CRC-B18 F 51 III S-colon AC 1.2 CRC-B19 M 67 IV Rectum AC 16.2 CRC-B20 M 71 I Rectum AC 83.7 CRC-B21 M 59 I S-colon AC 3.0 CRC-B22 F 66 IV S-colon AC 18.5 CRC-B23 F 78 IV A-colon AC 12.6 CRC-B24 M 55 III A-colon AC 1.2 CRC-B25 M 62 III Rectum AC 2.5 CRC-B26 M 38 III Rectum AC 6.1 CRC-B27 F 65 III D-colon AC 3.5 CRC-B28 M 49 III S-colon, T-colon AC 3.8 CRC-B29 F 59 II A-colon AC One CRC-B30 M 62 II S-colon AC - CRC-B31 F 54 IV S-colon AC 27.9 CRC-B32 M 66 III S-colon AC 10.7 CRC-B33 M 84 II S-colon AC 11.3 CRC-B34 F 54 III S-colon AC 8.8 CRC-B35 M 68 II Rectum AC 5.8 CRC-B36 M 54 II A-colon AC 1.1 CRC-B37 M 62 III S-colon AC 10.8 CRC-B38 F 60 III S-colon, A-colon AC 28.5 CRC-B39 F 51 II D-colon AC 5.9 CRC-B40 F 73 III Rectum AC 3.7 CRC-B41 F 54 III D-colon AC 1122.2 CRC-B42 M 64 I Rectum AC 2.5 CRC-B43 F 69 II S-colon, A-colon AC 5.1 CRC-B44 M 39 II S-colon AC 2.9 CRC-B45 M 74 II Rectum AC 7.9 CRC-B46 F 59 III Rectum AC 1.4 CRC-B47 M 56 I Rectum AC 2.6 CRC-B48 M 69 II Rectum AC 14.0 CRC-B49 M 58 II Rectum AC 10.2 CRC-B50 F 75 II Rectum AC 2.4 CRC-B51 M 47 II Rectum AC 3.2 CRC-B52 F 68 II Rectum AC 0.7 CRC-B53 M 52 III Rectum AC 2.9 CRC-B54 M 68 I Rectum AC 7.0 CRC-B55 M 51 II Rectum AC 1.4 CRC-B56 M 66 0 Rectum AC 1.2 CRC-B57 M 74 0 Rectum AC 4.5 CRC-B58 M 43 II Rectum AC 12.3 CRC-B59 M 68 III Rectum AC 2.5 CRC-B60 M 68 III Rectum AC 19.4 CRC-B61 F 56 I Rectum AC 2.3 CRC-B62 M 63 0 Rectum AC 1.3 CRC-B63 M 65 II Rectum AC 2.1 CRC-B64 M 60 II Rectum AC 4.6 CRC-B65 M 51 II Rectum AC 1.3 CRC-B66 M 44 0 Rectum AC 2.2 CRC-B67 M 61 II Rectum AC 2.0 CRC-B68 M 57 III Rectum AC 2.2 CRC-B69 M 41 II Rectum AC 3.1 CRC-B70 M 50 I Rectum AC 4.9 CRC-B71 F 56 III Rectum AC 1.0 CRC-B72 M 54 III Rectum AC 1.7 CRC-B73 F 69 I Rectum AC 1.5 CRC-B74 M 54 I Rectum AC 2.6 CRC-B75 M 61 II Rectum AC 3.7 CRC-B76 M 72 III Rectum AC 3.0 CRC-B77 F 71 III Rectum AC 1.8 CRC-B78 M 54 II Rectum AC 3.0 CRC-B79 M 77 II Rectum AC 1.6 CRC-B80 M 67 III Rectum AC 1.1 CRC-B81 M 59 II Rectum AC 7.2 CRC-B82 M 56 III Rectum AC 9.0 CRC-B83 F 51 I Rectum AC 1.5 CRC-B84 F 67 III Rectum AC 3.4 CRC-B85 F 76 III Rectum AC 1.0 CRC-B86 F 38 III Rectum AC 0.7 CRC-B87 M 53 II Rectum AC 3.3 CRC-B88 M 58 III Rectum AC 1.6 CRC-B89 M 69 III Rectum AC 6.4 CRC-B90 F 60 I Rectum AC 1.2 CRC-B91 M 52 II Rectum AC 4.0 CRC-B92 M 59 III Rectum AC 2.8 CRC-B93 F 56 III Rectum AC 2.3 CRC-B94 F 68 I Rectum AC 2.0 CRC-B95 M 65 I Rectum AC 1.6 CRC-B96 M 33 II Rectum AC 1.9 CRC-B97 M 61 III Rectum AC 3.2 CRC-B98 F 41 III Rectum AC 1.5 CRC-B99 M 61 I Rectum AC 1.6 CRC-B100 F 34 III Rectum AC 5.2 CRC-B101 M 47 III Rectum AC 2.3 CRC-B102 F 61 III A-colon AC 30.4 CRC-B103 M 71 IV A-colon AC 33.5 CRC-B104 M 44 III A-colon MAC 77.6 CRC-B105 F 71 III Rectum AC 1.0 CRC-B106 M 59 II S-colon AC 2.9 CRC-B107 M 79 IV Rectum AC 4.1 CRC-B108 M 66 II S-colon AC 1.7 CRC-B109 M 78 III S-colon AC 2.5 CRC-B110 F 53 II S-colon AC 1.3 CRC-B111 M 50 III S-colon AC 1.7 CRC-B112 F 77 II S-colon AC 2.2 CRC-B113 M 53 II S-colon ASC 8.6 CRC-B114 M 63 I Rectum AC 1.4 CRC-B115 F 71 III S-colon MAC 36.0 CRC-B116 F 79 II Rectum AC 6.2 CRC-B117 M 83 III A-colon MAC 10.9 CRC-B118 F 53 III A-colon AC 1.2 CRC-B119 F 72 III A-colon AC 2.8 CRC-B120 F 34 III D-colon AC 4.7 CRC-B121 M 70 III Rectum AC 3.5 CRC-B122 F 62 III A-colon AC 2.8 CRC-B123 M 45 II T-colon AC 1.8 CRC-B124 F 84 II A-colon AC 2.8 CRC-B125 M 74 III Rectum AC 11.1 CRC-B126 M 65 III D-colon AC 8.2 CRC-B127 M 69 II A-colon AC 20.1 CRC-B128 M 73 II A-colon AC 2.3 CRC-B129 M 61 II S-colon AC 2.1 CRC-B130 F 71 II S-colon AC 15.3 CRC-B131 F 56 I S-colon AC 0.7 CRC-B132 F 70 II S-colon AC 1.4 CRC-B133 F 62 III Rectum AC 235.4 CRC-B134 M 61 III S-colon AC 11.2 CRC-B135 F 52 III S-colon AC 6.4 CRC-B136 M 62 II S-colon AC 4.9 CRC-B137 F 61 III T-colon AC 13.9 CRC-B138 F 88 II A-colon AC 3.0 CRC-B139 M 73 II S-colon AC 16.5 CRC-B140 M 69 III A-colon AC 1.7 CRC-B141 M 71 III A-colon MAC 2.4 CRC-B142 F 45 0 Rectum AC - CRC-B143 M 66 0 Rectum AC 58.4

Control Sex Age
year CEA
ng / mL Control Sex Age
year CEA
ng / mL CONT-B1 M 64 2.4 CONT-B26 F 56 1.7 CONT-B2 M 53 3.3 CONT-B27 F 52 1.7 CONT-B3 M 63 0.9 CONT-B28 M 63 1.0 CONT-B4 M 57 1.5 CONT-B29 F 60 1.8 CONT-B5 F 66 1.6 CONT-B30 F 58 0.7 CONT-B6 M 60 1.5 CONT-B31 M 65 4.1 CONT-B7 M 57 2.2 CONT-B32 M 52 2.2 CONT-B8 M 53 1.9 CONT-B33 F 58 3.1 CONT-B9 M 60 0.8 CONT-B34 M 65 2.8 CONT-B10 F 50 2.6 CONT-B35 M 66 0.8 CONT-B11 F 53 2.6 CONT-B36 M 69 2.1 CONT-B12 M 64 1.5 CONT-B37 F 54 1.6 CONT-B13 F 60 3.7 CONT-B38 M 50 1.9 CONT-B14 M 58 1.2 CONT-B39 F 60 1.1 CONT-B15 F 66 1.1 CONT-B40 F 55 8.8 CONT-B16 M 57 2.9 CONT-B41 M 62 0.9 CONT-B17 F 68 5.5 CONT-B42 F 51 2.0 CONT-B18 M 56 1.7 CONT-B43 M 65 2.3 CONT-B19 M 51 3.4 CONT-B44 M 52 2.4 CONT-B20 F 56 1.3 CONT-B45 F 64 1.7 CONT-B21 M 57 1.5 CONT-B46 M 57 0.8 CONT-B22 M 61 4.2 CONT-B47 F 54 <0.5 CONT-B23 F 51 1.9 CONT-B48 F 59 0.8 CONT-B24 F 51 1.6 CONT-B49 F 65 1.6 CONT-B25 F 52 1.4 CONT-B50 F 68 1.6

BRC Sex Age
year Node ER ER% PR PR% HER2 Tumor Size
cm BRC-B1 F 45 ypN0 0 0% 0 0% 0 0.9 BRC-B2 F 59 pN0 0 0% 0 0% 3 1.1 BRC-B3 F 43 pN1 0 0% 0 0% 0 1.5 BRC-B4 F 46 pN1 8 100% 8 100% 0 1.3 BRC-B5 F 48 pN0 6 50-
60% 5 10-
20% 3 1.3 BRC-B6 F 39 pN0 0 0% 0 0% 0 2.2 BRC-B7 F 66 pN0 8 95% 8 95% 0 1.7 BRC-B8 F 39 ypN0 0 0% 0 0% 0 DCIS BRC-B9 F 37 pN0 7 70-
80% 8 80% 3 1.5 BRC-B10 F 64 pN0 8 95% 8 95% 0 0.5 BRC-B11 F 44 ypN1 7 90% 8 95% 0 2 BRC-B12 F 50 pN2 8 95% 8 100% 0 1.1 BRC-B13 F 47 pN0 7 70% 7 50-
60% One 0.5 BRC-B14 F 44 pN1 8 90% 8 95% One 0.6 BRC-B15 F 50 pN0 0 0% 0 0% 2 2.2 BRC-B16 F 53 pN0 7 95% 8 95% 0 1.1 BRC-B17 F 65 pN0 8 95% 7 40% 0 1.5 BRC-B18 F 39 pN1 7 >
95% 3 <
10% 0 2.2 BRC-B19 F 54 pN0 (i +) 7 95% 5 10-
30% One 1.7 BRC-B20 F 48 pN3a 7 90% 8 90% 0 3.2 BRC-B21 F 54 pN0 0 0% 0 0% 0 3 BRC-B22 F 43 pN0 7 50-
60% 7 50-
60% 3 2.3 BRC-B23 F 61 pN0 8 95% 8 95% 0 1.6 BRC-B24 F 54 - 0 0% 0 0% 3 - BRC-B25 F 46 pN0 7 80% 8 95% 0 2.2

NHL Sex Age
year Stage Involved Site Subtype IPI NHL-B1 M 58 3 multiple DLBL 2 NHL-B2 F 24 4 multiple DLBL 4 NHL-B3 M 56 4 multiple DLBL 3 NHL-B4 M 54 2 stomach DLBL One NHL-B5 F 76 4 multiple DLBL 3 NHL-B6 F 69 4 multiple Mantle cell L 4 NHL-B7 F 49 One mandibular area DLBL 0 NHL-B8 F 64 4 multiple DLBL 5 NHL-B9 M 44 4 multiple DLBL 2 NHL-B10 F 70 4 multiple DLBL 3 NHL-B11 M 25 4 multiple NK / T cell L 3 NHL-B12 F 48 One neck, submandibular DLBL 0 NHL-B13 F 48 One breast DLBL One NHL-B14 M 48 4 multiple DLBL 3 NHL-B15 M 67 4 multiple MZBCL 2

GC Sex Age
year CEA
ng / mL Stage GC Sex Age
year CEA
ng / mL Stage GC-B1 F 62 8.30 III GC-B29 F 57 - II GC-B2 M 64 - III GC-B30 M 78 - II GC-B3 M 58 6.89 III GC-B31 M 75 - II GC-B4 M 34 3.57 IV GC-B32 F 67 - II GC-B5 M 69 1.39 IV GC-B33 M 50 - II GC-B6 M 49 1.67 IV GC-B34 F 60 - II GC-B7 F 34 13.44 IV GC-B35 F 47 - II GC-B8 F 47 2.86 III GC-B36 M 69 - II GC-B9 F 40 0.64 IV GC-B37 M 72 - II GC-B10 F 60 1.20 II GC-B38 F 49 - II GC-B11 M 66 11.68 IV GC-B39 F 55 - III GC-B12 M 73 - II GC-B40 F 46 - III GC-B13 M 53 1.00 III GC-B41 M 64 - III GC-B14 M 51 5.60 IV GC-B42 M 53 - III GC-B15 M 42 0.69 III GC-B43 M 61 - III GC-B16 M 81 - III GC-B44 F 81 - III GC-B17 M 72 - II GC-B45 F 36 - III GC-B18 M 66 1.22 IV GC-B46 M 50 - IV GC-B19 F 49 - II GC-B47 M 55 - IV GC-B20 M 48 - I GC-B48 M 66 - IV GC-B21 M 51 - I GC-B49 F 40 - IV GC-B22 M 44 - I GC-B50 M 61 - IV GC-B23 F 44 - I GC-B51 M 70 - IV GC-B24 M 61 - I GC-B52 M 39 - IV GC-B25 M 76 - I GC-B53 M 70 - IV GC-B26 M 51 - I GC-B54 F 71 - IV GC-B27 F 40 - I GC-B55 F 52 - IV GC-B28 M 62 I

OVC Age
year Stage Histology OVC-B1 56 IIIc Clear cell carcinoma OVC-B2 52 IIa Endometrioid adenocarcinoma OVC-B3 63 IV Papillary serous adenocarcinoma OVC-B4 55 Ia Malignant Brenner tumor OVC-B5 47 IIIc Papillary serous adenocarcinoma OVC-B6 50 Ic Clear cell carcinoma OVC-B7 68 Ib Serous adenocarcinoma OVC-B8 74 IIIc Papillary serous adenocarcinoma OVC-B9 43 Ic Mucinous adenocarcinoma OVC-B10 44 IIIc Papillary serous adenocarcinoma OVC-B11 54 IIIc Papillary serous adenocarcinoma OVC-B12 55 IV Serous adenocarcinoma OVC-B13 72 IIIc Serous adenocarcinoma OVC-B14 58 IIIc Mucinous adenocarcinoma OVC-B15 44 IIIc Papillary serous adenocarcinoma OVC-B16 57 IV Serous adenocarcinoma OVC-B17 54 IIIc Papillary serous adenocarcinoma OVC-B18 73 IIIc Serous adenocarcinoma OVC-B19 47 IIIc Papillary serous adenocarcinoma OVC-B20 40 Ic Papillary serous adenocarcinoma OVC-B21 74 IIb Transitional cell carcinoma OVC-B22 65 IIIc Papillary serous adenocarcinoma OVC-B23 47 IV Serous adenocarcinoma OVC-B24 58 IIc Serous adenocarcinoma OVC-B25 57 Ib Mixed cell adenocarcinoma

TA Sex Age
year FOBT CEA
ng / mL TA-B1 M 70 Negative 2.0 TA-B2 M 58 Negative 2.0 TA-B3 M 52 Negative 0.6 TA-B4 M 48 - ^a 2.6 TA-B5 F 59 Negative 5.8 TA-B6 M 77 Negative 5.1 TA-B7 F 53 Negative 3.2 TA-B8 M 63 Negative - TA-B9 F 68 - - TA-B10 M 54 Negative 79.4 TA-B11 M 69 - 1.1 TA-B12 M 56 - 0.5 TA-B13 M 53 - 2.0 TA-B14 F 76 Negative 2.7 TA-B15 M 63 Positive 3.2 TA-B16 F 54 Negative 0.7 TA-B17 F 62 - 1.1 TA-B18 F 64 Negative 1.6 TA-B19 F 46 Positive 1.2

2-2. Sample Preparation - Serum Preparation and Mass Spectrometry

Four volumes of methanol / chloroform (2: 1, v / v) were vigorously mixed with 25 μl of serum and then incubated at room temperature for 10 minutes. The mixture was centrifuged at 6000 x g for 10 minutes at 4 &lt; 0 &gt; C. The resulting supernatant was completely dried in a concentrator for 1 hour and dissolved in 30 μl of 50% acetonitrile / 0.1% trifluoroacetic acid (TFA) in a vortexer for 30 minutes .

The methanol / chloroform extract was mixed (1:12, v / v) with a solution of alpha-cyano-4-hydroxycinnamic acid in 50% acetonitrile / 0.1% 1 [mu] l of the mixture was placed on a MALDI-target plate. The mass spectra of CRC patients and non-patient serum extracts were measured using a Proteomics Analyzer (Applied Biosystems, Foster City, CA, USA).

For one sample, mass spectral data is extracted as the mean of the 20 repeatedly measured spectra. The mass value interval of all the individual samples was adjusted so that the maximum mass value was about 2500 m / z. To minimize experimental errors, various factors were examined including focus mass, laser intensity, target plate, and data acquisition time.

500 m / z and 5000 were used with fixed focus mass and laser intensity, respectively. All samples were measured at least five times under different extraction and different data acquisition times, with fixed focus mass and laser intensity. In the case of the set A _{1 in which} the weight per weight ion was calculated, an additional measurement was performed.

Accordingly, the low mass ion detector 1000 can extract mass spectra of low mass ions from the serum sample through the above-described process using a MALDI-TOF mass spectrometer.

2-3. Discrimination strategy

In order for the discriminant formula to be CRC-specific, CRC patients should be distinguished from other cancer patients as well as the normal control group by discrimination. In this example, we would like to use BRC patients, NHL patients, and GC patients as other cancer patients. Table 20 shows the results of applying the linear discriminant analysis method based on the principal component analysis to discriminate between the CRC patient group and the non-patient group (normal control group, BRC group, NHL patient group and GC patient group) as one discrimination formula. In particular, the specificity of the normal control group is as low as 69.28%. From this, one discriminant can not distinguish between CRC patients and non-CR patients.

Sensitivity 99.25% Set A ₁ True CRC True Non-CRC

Specificity

CONT 69.28% CONT BRC NHL
GC BRC 97.30% Predicted CRC 132 47 3 One 0 NHL 97.22% Predicted Non-CRC One 106 108 35 29 GC 100.0%

As shown in FIG. 2 and Table 4, excellent discrimination results were obtained for the CRC patient group and the normal control group, so that the CRC patient group and the other cancer patient group could be distinguished from each other, and the results are shown in Table 21. The specificity of the NHL patient group was 83.33%, which is relatively low compared to other cancer patients.

Sensitivity 99.25% Set A ₁ True CRC True Non-CRC

Specificity

BRC 99.10% BRC NHL
GC NHL 83.33% Predicted CRC 132 One 6 One GC 96.55% Predicted Non-CRC One 110 30 28 Total 95.45%

Therefore, the discrimination between CRC patients and non-CRC patients was discriminated by CRC by using two discriminant equations using the first discriminant, which distinguishes between CRC patients and normal controls, and the second discriminant, which distinguishes CRC patients from other cancer patients The CRC patient is judged as a CRC patient, and in the case where it is discriminated as a CRC non-patient by any of the two discrimination equations, it can be judged as a CRC non-patient.

2-4. First training set A ₀ Selection of weight and calculation of weight by mass ion

The results of discrimination shown in Table 4 and Table 21 are excellent, but the sensitivity and the specificity are not all 100%. In the present invention, a first training set A ₀ having sensitivity and specificity of a predetermined value is selected, and a weight-by-weight of each of the first training set A ₀ is calculated. In this embodiment, Both sensitivity and specificity are 100%.

A method of selecting a first training set A ₀ having sensitivity and specificity of a predetermined value will be described with reference to FIG.

The first discrimination score calculating means 2200 arranges and imports the low mass ion mass spectra of the CRC patient group and the normal control group in the set A ₁ (D111) through the steps D111 to D114 shown in FIG. 10 (D111) Strengths were normalized (D112), normalized peak intensities were Pareto-scaled (D113), and Pareto-scaled peak intensities were subjected to biostatistical analysis (D114).

The discriminant score can be calculated by selecting one of various biostatistical analysis methods. In this embodiment, linear discriminant analysis based on principal component analysis is performed. Sensitivity and specificity were calculated through discriminant scores (D115), and the results are shown in FIG. 2 and Table 4 described above.

Next, except for the (D116), the sensitivity or specificity case is less than a respective threshold value false positives (false positive) or false negative (false negative) case by setting the threshold value CN ₁ and the specific threshold value CN ₂ in view of the sensitivity (D117).

In the present embodiment, by setting both the sensitivity threshold CN ₁ and the threshold value CN ₂ of the specificity as 1, a first training set A ₀ with a sensitivity of 100% and a specificity of 100% is found. That is, the two false positive cases and the two false positive cases shown in Table 4 were excluded, and the excluded sets were again subjected to the steps D111 to D115. The sensitivity and specificity were not 100% even after the steps D111 to D115 were repeated for the excluded assemblies. By repeating the steps D111 to D117 several times, the first training set A having sensitivity and specificity of 100% ₀ was found (D118).

In the case of the first discriminant, which distinguishes between the CRC patient group and the normal control group, there are 5 false negatives, 10 false negatives, 8 false negatives, and 9 false positive cases, After excluding the 1 false positive cases (1 BRC, 8 NHL, and 1 GC), we reached the discrimination result that both the sensitivity and the specificity were both 100%. Through this process, it is possible to derive the mass-ion-specific factor load value which gives the discrimination result that both the sensitivity and the specificity are 100% (D119).

The above-described series of processes may be performed through the parameter set calculation means 2300 including the first training set selection means 2310.

2-5. Application of discriminant equation

The process of applying the discriminant equation to the discriminant sample will be described as follows.

First of all, MarkerView ^TM has features that can be used for similar purposes. That is, the linear discriminant analysis based on principal component analysis can be applied to only some of the sample data imported together, and the remaining samples can be discriminated by the discriminant thus constructed. With this function, the first training set and the discrimination sample are imported together, and then only the first training set is selected, and the linear discrimination analysis based on the principal component analysis is performed to know how the discrimination target sample is judged.

However, since the peak alignment is performed during the import process of MarkerView (TM), since there is no function of aligning the peaks of the sample to be discriminated in accordance with the first training set, the peak table peak table, m / z column and sample peak intensity column) and the first training set part of the peak table generated when the first training set and the discrimination sample are imported together are not the same. The peak intensity matrix portions are also different, and the m / z values corresponding to the same peak intensity row are not always the same. Therefore, in order to calculate the discrimination score by applying the discrimination equation made in the first training set to the discrimination sample, the peak table generated when the first training set and the discrimination sample are imported together is used as the peak when the first training set is imported The table should be preceded by realignment.

When multiple samples to be discriminated are imported together with the first training set, the degree of misalignment becomes worse. Therefore, in this embodiment, one discrimination sample is added to the first training set for all the discrimination samples, and then the rearranging, normalization, and pareto scaling are performed.

This will be described in more detail with reference to FIG.

First, a step of sorting and importing low mass ion mass spectra of a discrimination target sample to a first training set is performed (D211).

However, in the present embodiment, since the MarkerView ^TM does not have a function of sorting and importing the discrimination target sample into the first training set, the low mass ion mass spectrum of the discrimination sample is imported together with the first training set A program for realigning the peak table to the peak table created when only the first training set was imported was created to extract the low mass ion mass spectrum of the discriminated sample arranged in the first training set. However, it is more preferable to arrange and import the discriminant sample from the beginning into the first training set without rearranging it, and this is also possible to create and implement the program.

Next, the imported peak intensities are normalized (D212) and the normalized peak intensities are Pareto-scaled (D213).

Next, the discrimination score was calculated from the peak intensity of the Pareto-scaled low-mass ions and the factor loadings of mass ions obtained by the linear discriminant analysis based on principal component analysis on the first training set (D214).

It is determined whether the thus calculated discrimination score is greater than the reference value CS (D215). If it is greater than the reference value CS, it is determined to be positive (D216). In this embodiment, the reference value CS is preferably zero.

The above-described series of steps may be performed through the second alignment means 2500, the second discrimination score calculation means 2600, and the colon cancer determination means 2700.

The time to configure the first training set A ₀₁ in the set A ₁ to the discriminant one kinds of negative eight CRC patients and nine normal control samples and the first training set from set A ₁ to the discriminant two kinds were A _{For each of} the five CRC patients, one BRC patient, eight NHL patients, and one GC patient sample that were excluded when constructing the ₀₂ , calculate the discrimination score by applying the mass-specific factor loadings calculated in Section 2-4 above saw. Because it was already excluded when constructing the first training set A ₀₁ and A ₀₂ , it was expected to be judged as a false positive or false negative case, and the calculation result was judged as false positive or false negative case as expected. 12 and 13 show the result of discriminating the set A ₁ by applying the mass load factor values calculated in the above section 2-4. FIG. 12 shows the first type discriminant, FIG. 13 shows the second kind discriminant The results of the equation are shown.

2-6. Composition of the preliminary discrimination formula

In the above-described conventional technique, the discrimination score is calculated using all the mass ions considered in the linear discriminant analysis based on the principal component analysis, and then the CRC patient is judged according to the discrimination score. In the present invention, however, the discriminant expressing robust discrimination performance is derived We made a preliminary discrimination equation using only mass ions which make a big contribution to the discrimination score. The "preliminary discrimination formula" refers to a discriminant function as an intermediate step for deriving a discriminant to be ultimately obtained according to the present invention. The low-mass ions constituting the discriminant form a "reserve " Candidate ".

Through the procedure of the flowchart shown in FIG. 14, predetermined mass ions having a large influence on the discrimination score among 10,000 mass ions were selected. In the present embodiment, the number of predetermined mass ions was 278 in the case of the first kind discrimination formula, and 383 in the case of the second kind discrimination formula.

As described above, the maximum number of peaks in the import condition was 10,000, and a sufficient number of samples were imported together. As a result, the discriminant formed by the linear discriminant analysis based on the principal component analysis of MarkerView ^TM is composed of 10,000 terms. However, since 10,000 terms would not be of equal importance in distinguishing between CRC patients and non-patients, mass ions having a large influence on the discrimination score among the 10,000 mass ions were selected in two steps according to the procedure of FIG. This step can be regarded as a process of removing unnecessary mass ions in discriminating between CRC patient and non-patient among 10,000 mass ions.

From 10,000 term values the absolute value of a product of a peak intensity and a mass ion-specific factor loading values were primarily selected by the large mass ions than the threshold value CT ₁ on a case-by-case basis (D121). The threshold value CT ₁ in this embodiment is preferably 0.1.

Next, among the mass ions selected on a case-by-case basis, among the entire cases of the first training set, mass ions commonly appearing in cases of ₂ % or more of the threshold value CT were secondarily selected (D122). In this embodiment, the threshold value CT ₂ is preferably 50. In other words, the preliminary discriminant is composed of only the mass ions that appear in common in at least 135 cases among the 269 cases of the first training set.

The discrimination score was again calculated using only selected mass ions and the sensitivity and specificity were calculated (D123). Again, by setting the threshold value CN ₃ and specificity threshold CN ₄ of the sensitivity (D124), the sensitivity or specificity is less than the respective threshold value in the threshold value CT ₁ and / or D122 step in D121 step by changing the threshold value CT ₂ using (D125) D121 to D124 was repeated steps: In the present embodiment, it is preferable that the sensitivity threshold CN ₃ and the threshold value CN ₄ of the specificity are 0.9, respectively.

The preliminary candidates of the low mass ions for the CRC diagnosis were constituted by the mass ions selected through the above steps (D126). In this embodiment, 278 of the 10,000 kinds of mass ions are classified into the first type discrimination type, 383 mass ions were selected. Table 22 and Table 23 show the result of discrimination of the first training set A ₀₁ and A ₀₂ in the first type preliminary discrimination type and the second type discrimination type. The discrimination performance such as the sensitivity and the specificity was slightly lowered from 100% Although it is calculated by using only mass ions not more than 3 to 4% of the total number of mass ions, it can be confirmed that there is an excellent discrimination result which is comparable to that when using all mass ions.

15 and 16 show the result of discrimination of the set A ₁ in the preliminary discrimination formula. FIG. 15 shows the results of the first type discrimination method and FIG. 16 shows the results of the second type discrimination method. It can be seen that the number of mass ions used in the calculation is sharply reduced, whereas the range of the discrimination score is not so. From this, it can be seen that 10,000 mass ions are not necessary in distinguishing between CRC patients and non-patients have.

Sensitivity 99.20% Set A ₀₁ True CRC True CONT Specificity 97.92% Predicted CRC 124 3 PPV 97.64% Predicted CONT One 141 NPV 99.30%

Set A ₀₂ True CRC True Non-CRC
Sensitivity 98.44% BRC NHL
GC Specificity 98.19% Predicted CRC 126 2 One 0 PPV 97.67% Predicted Non-CRC 2 108 27 28 NPV 98.79%

The above-described series of processes can be effectively performed through the colon cancer diagnostic ion selecting means 2400 including the candidate ion set selection means 2410.

2-7. Composition of the final discriminant equation

In the process of constructing the preliminary discriminant, mass ions which contribute a great numerical contribution to the discrimination score among the 10,000 imported mass ions were extracted. However, among the mass ions, there was no problem in the first training set A ₀ , but in the discrimination of the mass spectra re-measured for the serum of the same CRC patient group and the non-patient group, or in the discrimination for the new CRC patient group and the non- Since the mass ions that may degrade the discrimination performance are also included, it is necessary to actively remove them. In the final discrimination process, the low-mass ions for CRC diagnosis are finally determined through this step.

In order to verify the robustness of the discriminant equation, five experiments were repeatedly performed on the set A ₁ , and five independent experiments were performed on the independent and independent sets A ₂ and B, respectively. In the repeated measurement of the mass spectrum, a process of melting and dissolving the serum described above and a process of mixing the serum with methanol / chloroform are newly introduced, and vaporization, desorption, and the like using a laser beam are performed. And ionization process can not be said to be the same in repetitive experiments, and disturbances from various causes that have not yet been revealed can be intervened. Therefore, it can not be excluded that a certain deviation occurs in the discrimination score for the repeated mass spectra. Therefore, in this embodiment, the average discrimination score was calculated for the samples repeated five times and the judgment was made.

Table 24 shows the results of discrimination of sets A and B by the discriminant of 10,000 terms, which is the result of the linear discriminant analysis based on the principal component analysis of the prior art MarkerView ^TM . Table 25 shows the results of the first type discriminant equation having 278 terms and 383 terms The results of discrimination between sets A and B are shown in the second kind discriminant equation. In the table, CRC LOME (colorectal cancer low mass discriminant equation) 1 is the first discriminant, CRC LOME 2 is the second discriminant, and the number following it is the number of low mass ions it means. Table 26 shows the discrimination performance for set B, which is a verification set. The numbers in parentheses are the discrimination performance when the TA patient group is included in the CRC patient group. TA patients are more likely to develop into CRC patients, and discrimination of TA patients as CRC patients is a desirable result for the purpose of early detection.

CRC LOME 1-10000 CRC LOME 1-10000 Sete True
CRC True Non-CRC Set B True CRC True Non-CRC CONT BRC NHL GC CONT BRC NHL GC OVC TA Predicted CRC 131 18 26 36 36 Predicted CRC 68 0 7 4 23 6 13 Predicted Non-CRC 146 185 110 15 50 Predicted Non-CRC 75 50 18 11 32 19 6 CRC LOME 2-10000 CRC LOME 2-10000 Sete True
CRC True Non-CRC Set B True CRC True Non-CRC CONT BRC NHL GC CONT BRC NHL GC OVC TA Predicted CRC 109 120 21 3 2 Predicted CRC 43 0 4 0 0 0 9 Predicted Non-CRC 168 83 115 48 84 Predicted Non-CRC 100 50 21 15 55 25 10 CRC LOMEs 1 & 2 CRC LOMEs 1 & 2 Sete True
CRC True Non-CRC Set B True CRC True Non-CRC CONT BRC NHL GC CONT BRC NHL GC OVC TA Predicted CRC 80 17 3 3 One Predicted CRC 39 0 One 0 0 0 8 Predicted Non-CRC 197 186 133 48 85 Predicted Non-CRC 104 50 24 15 55 25 11

CRC LOME 1-278 CRC LOME 1-278 Sete True
CRC True Non-CRC Set B True CRC True Non-CRC CONT BRC NHL GC CONT BRC NHL GC OVC TA Predicted CRC 136 14 26 36 35 Predicted CRC 70 0 5 4 24 7 13 Predicted Non-CRC 141 189 110 15 51 Predicted Non-CRC 73 50 20 11 31 18 6
CRC LOME 2-383 CRC LOME 2-383 Sete True
CRC True Non-CRC Set B True CRC True Non-CRC CONT BRC NHL GC CONT BRC NHL GC OVC TA Predicted CRC 106 121 21 3 4 Predicted CRC 43 0 3 0 0 0 9 Predicted Non-CRC 171 82 115 48 82 Predicted Non-CRC 100 50 22 15 55 25 10 CRC LOMEs 1 & 2 CRC LOMEs 1 & 2 Sete True
CRC True Non-CRC Set B True CRC True Non-CRC CONT BRC NHL GC CONT BRC NHL GC OVC TA Predicted CRC 75 13 3 3 One Predicted CRC 39 0 One 0 0 0 8 Predicted Non-CRC 202 190 133 48 85 Predicted Non-CRC 104 50 24 15 55 25 11

Set B Sensitivity (%) Specificity (%) PPV (%) NPV (%) CRC LOME 1-10000 &
CRC LOME 2-10000 27.27 (29.01) 95.24 (99.41) 81.25 (97.92) 63.38 (59.51) CRC LOME 1-278 &
CRC LOME 2-383 27.27 (29.01) 95.24 (99.41) 81.25 (97.92) 63.38 (59.51) CRC LOME 1-104 &
CRC LOME 2-23 94.41 (93.21) 88.36 (96.47) 85.99 (96.18) 95.43 (93.71)

Table 26 shows that the discriminant composed of 10,000 mass ions showed a particularly low sensitivity in the set B, although the complete discrimination performance in the first training set A ₀ was shown. The results of discrimination in group B were unsatisfactory despite the excellent discrimination performance (Table 22, Table 23) in the first training group A _{0 and} the first training class A 1 of the first type and the second kind.

Therefore, in a preferred embodiment of the present invention, the steps of the flowchart shown in FIG. 17 have been performed in order to improve the preliminary discrimination equation to a robust discriminant.

First, the mass ions of the preliminary candidate group are classified into a set of high sensitivity and a high specificity group (D131). Here, mass ions in the hypertonic set are mass ions whose mass ion specificity is higher than specificity, and mass ions in the high specificity set are the opposite.

Next, the mass ions of the hypertonic set and the mass ions of the high specificity set are sorted in descending order of the sensitivity and specificity of each mass ion, respectively, and then {Sns ₁ , Sns ₂ , Sns ₃ ... Sns _I } {Spc ₁ , Spc ₂ , Spc ₃ ... Spc _J }, each of the top two mass ions {Sns ₁ , Sns ₂ , Spc ₁ , Spc ₂ }, combinations (11) that can consist of two or more mass ions of four mass ions, (D132) were selected as the biomarker group.

Here, whether or not the combination of the highest performance can be objectively and universally selected according to the following criteria. The following criteria are in order of importance.

Criteria 1) The combination of sensitivity and specificity is higher.

Criteria 2) Combinations with fewer mass ions have higher performance.

Criterion 3) A combination with a large difference between the minimum discrimination score and the maximum discrimination score among the true cases is higher than the true discrimination case.

Next, the next one of the next higher mass ions {Sns ₃ , Spc ₃ } of each of the set of sensitivity sensitivity and the high specificity set is further taken, and the mass ions {Sns ₃ , the Spc _3} any set of four combining one or more of the {biomarker group}, {biomarker group , Sns 3}, {biomarker group, Spc 3}, set the highest performance of the _{{biomarker group, Sns 3, Spc} 3} Is selected again as a biomarker group (D133).

This process is repeated until there are no mass ions to add to both the hypertonic and high specificity sets (D134).

That is, when there are mass ions to be added to both the hyperhidimetic sensitivity set and the high specificity set, the above method (D133) is repeated. When no mass ion to be added to any of the hyperhidophilic set and the high specificity set is left , that of the remaining sets in the mass ion, and then the upper one mass ion {Sns _i or Spc _j} one more more taken as a combination of the mass ion {Sns _i or Spc _j} on the biomarker group and the biomarker group Biomarker group, biomarker group, Sns _i or Spc _j } among the two sets of biomarker groups.

This process is repeated until there is no remaining mass ion in the set of hypertonic and high specificity mass ions remaining, and when the biomarker group in which there is no remaining mass ion in both hypertonic and high specificity sets is bio Marker group 1 (CG) (D135).

Remove the biomarker group 1 (CG) from the preliminary candidate group (D136) and construct a set of high sensitivity and high specificity again with the remaining mass ions and repeat the above procedure. This process is repeated until either a set of critical sensitivity and a set of high specificity have less than two mass ions (D137).

The biomarker groups 1, 2, ... The CK biomarker groups are combined in order of accuracy to form the final biomarker group. Where the accuracy refers to the ratio of true positive and true negative cases in the overall case. In this embodiment, CK is preferably 1, 2 or 3 (D138).

Accordingly, the mass ions of the final biomarker group are determined as low mass ions for CRC diagnosis (D139).

Were in the set A _1, more accurately, selected pre-candidate group of the mass ion at the A ₀ set the portion, of the set A ₁ to when determining the final biomarker group from the pre-candidate group to address effectively the sum (overfitting) issues independent We expanded the training set by adding the set A ₂ .

As a result of performing the above procedures to distinguish between the CRC patient group and the normal control group, 104 low-mass ions for the first type CRC diagnosis were selected and samples for distinguishing between the CRC patient group and the other cancer group were performed RESULTS: Twenty-three of the low-mass ions for the second-type CRC diagnosis were selected. Mass values of low mass ions for the first and second species CRC diagnostics are shown in Tables 27 and 28. Such low-mass ions are referred to as " low-mass ions for the first kind of CRC diagnosis "and" low-mass ions for the second kind of CRC diagnosis &quot;, and finally the discriminant according to the present invention, Final discrimination formula for CRC diagnosis "and" second discrimination formula for CRC diagnosis for type II &quot;.

18.0260 190.0849 317.2311 401.0531 508.3407 548.2856 1206.5305 22.9797 191.0848 335.1862 423.0313 510.3265 566.8375 1207.5571 74.0948 191.3324 340.2241 428.1878 512.3119 581.1957 1465.6184 76.0763 195.0785 343.2451 454.2090 513.3177 582.1888 1466.6096 102.0916 212.3195 345.2583 465.3014 518.2931 583.2242 1467.5969 105.1078 231.0667 357.0666 466.1923 518.8555 656.0270 2450.9701 106.0899 235.0053 357.2784 468.1851 519.2967 667.3291 2451.9662 107.0477 256.0939 366.2310 469.2831 519.8598 709.3519 2452.9546 118.0822 266.9557 368.2551 477.1721 525.3449 710.3581 123.0395 267.9501 369.3302 478.1678 534.2739 711.3617 137.0423 288.2033 377.0570 480.1715 537.2800 712.3683 137.0729 291.0997 379.1438 482.3220 538.3306 713.3798 147.0573 295.0663 379.4765 483.3258 540.2629 991.6196 147.1058 300.1297 383.0529 496.8683 540.8144 992.6209 169.0653 301.1269 384.1745 497.7636 542.8457 1016.6113 181.0656 316.2288 388.2688 503.8719 544.8692 1020.4817

In Table 27, 1465.6184, 1466.6096, 1467.5969, 2450.9701, 2451.9662, 2452.9546 m / z were identified as the fibrinogen alpha chain and transthyretin.

60.0476 179.1451 280.2642 332.3224 486.6356 566.8737 138.0540 191.1277 281.1440 333.3324 488.6882 707.3475 172.6653 279.0855 296.2574 369.3406 544.8908 733.3569 173.1158 280.0895 312.3248 465.3161 551.3287

The above-described series of processes can be effectively performed through the colorectal cancer diagnosis ion selecting means 2400 including the candidate ion set selection means 2410 and the final ion set selection means 2420. [

2-8. Application and analysis of the final discriminant equation

Applying the final discriminant for first and second CRC diagnostics using the low-mass ions for the first kind and the second kind CRC diagnosis to the set B according to the method of the flowchart shown in FIG. 11, the judgment result can be obtained have.

The results of the final discrimination are shown in Figs. 18 and 19 and Tables 26 and 29. 18 and 19 show the results of judging by the average discrimination score of the five discrimination scores. FIG. 18 shows the result of the judgment of the set A and FIG. 19 shows the judgment result of the set B. FIG.

CRC LOME 1-104 CRC LOME 1-104 Sete True
CRC True Non-CRC Set B True CRC True Non-CRC CONT BRC NHL GC CONT BRC NHL GC OVC TA Predicted CRC 258 2 67 44 69 Predicted CRC 141 3 23 10 45 19 17 Predicted Non-CRC 19 201 69 7 17 Predicted Non-CRC 2 47 2 5 10 6 2
CRC LOME 2-23 CRC LOME 2-23 Sete True
CRC True Non-CRC Set B True CRC True Non-CRC CONT BRC NHL GC CONT BRC NHL GC OVC TA Predicted CRC 273 113 0 2 7 Predicted CRC 137 39 One 0 2 2 18 Predicted Non-CRC 4 90 136 49 79 Predicted Non-CRC 6 11 24 15 53 23 One CRC LOMEs 1 & 2 CRC LOMEs 1 & 2 Sete True
CRC True Non-CRC Set B True CRC True Non-CRC CONT BRC NHL GC CONT BRC NHL GC OVC TA Predicted CRC 254 0 0 2 5 Predicted CRC 135 One One 0 2 2 16 Predicted Non-CRC 23 203 136 49 81 Predicted Non-CRC 8 49 24 15 53 23 3

The sensitivity, specificity, positive predictive value, and negative predictive value of the set B, which is a verification set, are all above 85%. In addition, if the TA patient group is included in the CRC patient group, the discrimination performance is more than 90%, which indicates a very good discrimination result.

Table 30 shows the discrimination performance when the conventional FOBT analysis is performed among the analyzed samples and the discrimination performance of the present invention. The FOBT results for 96 samples from CRC patients and 49 samples from normal controls were 100% specific, but the sensitivity was only 50%.

In this example, the sensitivity of the commonly known FOBT is less than 60 ~ 85%. Even when compared with the discrimination performance of general FOBT, it can be seen that the present invention is comparable to FOBT in terms of specificity and shows excellent discrimination from the viewpoint of sensitivity, which proves the excellent discrimination performance of the present invention. Similar results are obtained in the training set. Table 31 shows the FOBT discrimination results for both the training set and the verification set and the discrimination results according to the present invention.

Set B CRC LOME 1-104 CRC LOME 1-104 &
CRC LOME 2-23 FOBT True
CRC True
CONT True
CRC True
CONT True
CRC True
CONT Predicted
CRC 94 3 91 One 48 0 Predicted
Non-CRC 2 46 5 48 48 49 Set B Sensitivity (%) Specificity (%) PPV (%) NPV (%) CRC LOME 1-104
97.92 93.88 96.91 95.83 CRC LOME 1-104 &
CRC LOME 2-23 94.79 97.96 98.91 90.57 FOBT 50.00 100.0 100.0 50.52

Control FOBT Prediction Control FOBT Prediction CRC LOME 1-104 CRC LOME 1-104 & CRC LOME 2-23 CRC LOME 1-104 CRC LOME 1-104 & CRC LOME 2-23 CONT-A1 - Non-CRC Non-CRC CONT-A103 Negative Non-CRC Non-CRC CONT-A2 Negative Non-CRC Non-CRC CONT-A104 - Non-CRC Non-CRC CONT-A3 Negative Non-CRC Non-CRC CONT-A105 Negative Non-CRC Non-CRC CONT-A4 Negative Non-CRC Non-CRC CONT-A106 - Non-CRC Non-CRC CONT-A5 Negative Non-CRC Non-CRC CONT-A107 Negative Non-CRC Non-CRC CONT-A6 Negative Non-CRC Non-CRC CONT-A108 Negative Non-CRC Non-CRC CONT-A7 Negative Non-CRC Non-CRC CONT-A109 Negative Non-CRC Non-CRC CONT-A8 Negative Non-CRC Non-CRC CONT-A110 Negative Non-CRC Non-CRC CONT-A9 Negative Non-CRC Non-CRC CONT-A111 Negative Non-CRC Non-CRC CONT-A10 Negative Non-CRC Non-CRC CONT-A112 Negative Non-CRC Non-CRC CONT-A11 Negative Non-CRC Non-CRC CONT-A113 Negative Non-CRC Non-CRC CONT-A12 Negative Non-CRC Non-CRC CONT-A114 Negative Non-CRC Non-CRC CONT-A13 Negative Non-CRC Non-CRC CONT-A115 Negative Non-CRC Non-CRC CONT-A14 Negative Non-CRC Non-CRC CONT-A116 Negative Non-CRC Non-CRC CONT-A15 Negative Non-CRC Non-CRC CONT-A117 Negative Non-CRC Non-CRC CONT-A16 Negative Non-CRC Non-CRC CONT-A118 Negative Non-CRC Non-CRC CONT-A17 Negative Non-CRC Non-CRC CONT-A119 Negative Non-CRC Non-CRC CONT-A18 Negative Non-CRC Non-CRC CONT-A120 - Non-CRC Non-CRC CONT-A19 - Non-CRC Non-CRC CONT-A121 - Non-CRC Non-CRC CONT-A20 Negative Non-CRC Non-CRC CONT-A122 Negative Non-CRC Non-CRC CONT-A21 Negative Non-CRC Non-CRC CONT-A123 Negative Non-CRC Non-CRC CONT-A22 Negative Non-CRC Non-CRC CONT-A124 Negative Non-CRC Non-CRC CONT-A23 Negative Non-CRC Non-CRC CONT-A125 Negative Non-CRC Non-CRC CONT-A24 Negative Non-CRC Non-CRC CONT-A126 Negative Non-CRC Non-CRC CONT-A25 Negative Non-CRC Non-CRC CONT-A127 Negative Non-CRC Non-CRC CONT-A26 Negative Non-CRC Non-CRC CONT-A128 Negative Non-CRC Non-CRC CONT-A27 Negative Non-CRC Non-CRC CONT-A129 Negative Non-CRC Non-CRC CONT-A28 Negative Non-CRC Non-CRC CONT-A130 - Non-CRC Non-CRC CONT-A29 Negative Non-CRC Non-CRC CONT-A131 Negative Non-CRC Non-CRC CONT-A30 Negative Non-CRC Non-CRC CONT-A132 Negative Non-CRC Non-CRC CONT-A31 Negative Non-CRC Non-CRC CONT-A133 Negative Non-CRC Non-CRC CONT-A32 Negative Non-CRC Non-CRC CONT-A134 - Non-CRC Non-CRC CONT-A33 Negative Non-CRC Non-CRC CONT-A135 - Non-CRC Non-CRC CONT-A34 Negative Non-CRC Non-CRC CONT-A136 Negative Non-CRC Non-CRC CONT-A35 Negative Non-CRC Non-CRC CONT-A137 Negative Non-CRC Non-CRC CONT-A36 Negative Non-CRC Non-CRC CONT-A138 Negative Non-CRC Non-CRC CONT-A37 Negative Non-CRC Non-CRC CONT-A139 Negative Non-CRC Non-CRC CONT-A38 Negative Non-CRC Non-CRC CONT-A140 Negative Non-CRC Non-CRC CONT-A39 Negative Non-CRC Non-CRC CONT-A141 Negative Non-CRC Non-CRC CONT-A40 - Non-CRC Non-CRC CONT-A142 Negative Non-CRC Non-CRC CONT-A41 Negative Non-CRC Non-CRC CONT-A143 Negative Non-CRC Non-CRC CONT-A42 Negative Non-CRC Non-CRC CONT-A144 Negative Non-CRC Non-CRC CONT-A43 Negative Non-CRC Non-CRC CONT-A145 Negative Non-CRC Non-CRC CONT-A44 - Non-CRC Non-CRC CONT-A146 - Non-CRC Non-CRC CONT-A45 Negative Non-CRC Non-CRC CONT-A147 - Non-CRC Non-CRC CONT-A46 Negative Non-CRC Non-CRC CONT-A148 Negative Non-CRC Non-CRC CONT-A47 Negative Non-CRC Non-CRC CONT-A149 - Non-CRC Non-CRC CONT-A48 Negative Non-CRC Non-CRC CONT-A150 Negative Non-CRC Non-CRC CONT-A49 Negative Non-CRC Non-CRC CONT-A151 Negative Non-CRC Non-CRC CONT-A50 Negative Non-CRC Non-CRC CONT-A152 Negative Non-CRC Non-CRC CONT-A51 Negative Non-CRC Non-CRC CONT-A153 - Non-CRC Non-CRC CONT-A52 Negative Non-CRC Non-CRC CONT-A154 Negative Non-CRC Non-CRC CONT-A53 Negative Non-CRC Non-CRC CONT-A155 Negative Non-CRC Non-CRC CONT-A54 Negative Non-CRC Non-CRC CONT-A156 Negative Non-CRC Non-CRC CONT-A55 Negative Non-CRC Non-CRC CONT-A157 Positive Non-CRC Non-CRC CONT-A56 Negative Non-CRC Non-CRC CONT-A158 Negative Non-CRC Non-CRC CONT-A57 Negative Non-CRC Non-CRC CONT-A159 Negative Non-CRC Non-CRC CONT-A58 Negative Non-CRC Non-CRC CONT-A160 Negative Non-CRC Non-CRC CONT-A59 Negative Non-CRC Non-CRC CONT-A161 Negative Non-CRC Non-CRC CONT-A60 Negative Non-CRC Non-CRC CONT-A162 Negative Non-CRC Non-CRC CONT-A61 Negative Non-CRC Non-CRC CONT-A163 Negative Non-CRC Non-CRC CONT-A62 Negative Non-CRC Non-CRC CONT-A164 - Non-CRC Non-CRC CONT-A63 Negative Non-CRC Non-CRC CONT-A165 Negative Non-CRC Non-CRC CONT-A64 Negative Non-CRC Non-CRC CONT-A166 Negative Non-CRC Non-CRC CONT-A65 Negative Non-CRC Non-CRC CONT-A167 Negative Non-CRC Non-CRC CONT-A66 Negative Non-CRC Non-CRC CONT-A168 Negative Non-CRC Non-CRC CONT-A67 Negative Non-CRC Non-CRC CONT-A169 Negative Non-CRC Non-CRC CONT-A68 Negative Non-CRC Non-CRC CONT-A170 Negative Non-CRC Non-CRC CONT-A69 Negative Non-CRC Non-CRC CONT-A171 Negative Non-CRC Non-CRC CONT-A70 Negative Non-CRC Non-CRC CONT-A172 Negative Non-CRC Non-CRC CONT-A71 Negative Non-CRC Non-CRC CONT-A173 Negative Non-CRC Non-CRC CONT-A72 Negative Non-CRC Non-CRC CONT-A174 Negative Non-CRC Non-CRC CONT-A73 Negative Non-CRC Non-CRC CONT-A175 Negative Non-CRC Non-CRC CONT-A74 Negative Non-CRC Non-CRC CONT-A176 - Non-CRC Non-CRC CONT-A75 Negative Non-CRC Non-CRC CONT-A177 Negative Non-CRC Non-CRC CONT-A76 Negative Non-CRC Non-CRC CONT-A178 Negative Non-CRC Non-CRC CONT-A77 Negative Non-CRC Non-CRC CONT-A179 Negative Non-CRC Non-CRC CONT-A78 Negative Non-CRC Non-CRC CONT-A180 Negative Non-CRC Non-CRC CONT-A79 Negative Non-CRC Non-CRC CONT-A181 Negative Non-CRC Non-CRC CONT-A80 Negative Non-CRC Non-CRC CONT-A182 Negative Non-CRC Non-CRC CONT-A81 Negative Non-CRC Non-CRC CONT-A183 Negative Non-CRC Non-CRC CONT-A82 Negative Non-CRC Non-CRC CONT-A184 Negative Non-CRC Non-CRC CONT-A83 Negative CRC Non-CRC CONT-A185 Negative Non-CRC Non-CRC CONT-A84 Negative Non-CRC Non-CRC CONT-A186 Negative Non-CRC Non-CRC CONT-A85 Negative Non-CRC Non-CRC CONT-A187 Negative Non-CRC Non-CRC CONT-A86 Negative Non-CRC Non-CRC CONT-A188 - Non-CRC Non-CRC CONT-A87 Negative Non-CRC Non-CRC CONT-A189 Negative Non-CRC Non-CRC CONT-A88 Negative Non-CRC Non-CRC CONT-A190 Negative Non-CRC Non-CRC CONT-A89 Negative Non-CRC Non-CRC CONT-A191 Negative Non-CRC Non-CRC CONT-A90 Negative Non-CRC Non-CRC CONT-A192 Negative Non-CRC Non-CRC CONT-A91 Negative Non-CRC Non-CRC CONT-A193 Negative Non-CRC Non-CRC CONT-A92 Negative Non-CRC Non-CRC CONT-A194 Negative Non-CRC Non-CRC CONT-A93 Negative Non-CRC Non-CRC CONT-A195 Negative Non-CRC Non-CRC CONT-A94 Negative Non-CRC Non-CRC CONT-A196 Negative Non-CRC Non-CRC CONT-A95 - Non-CRC Non-CRC CONT-A197 Negative Non-CRC Non-CRC CONT-A96 Negative Non-CRC Non-CRC CONT-A198 Negative Non-CRC Non-CRC CONT-A97 Negative Non-CRC Non-CRC CONT-A199 Negative Non-CRC Non-CRC CONT-A98 Negative Non-CRC Non-CRC CONT-A200 Negative Non-CRC Non-CRC CONT-A99 Negative Non-CRC Non-CRC CONT-A201 Negative CRC Non-CRC CONT-A100 Negative Non-CRC Non-CRC CONT-A202 Negative Non-CRC Non-CRC CONT-A101 Negative Non-CRC Non-CRC CONT-A203 Negative Non-CRC Non-CRC CONT-A102 Negative Non-CRC Non-CRC CONT-B1 Negative Non-CRC Non-CRC CONT-B26 Negative Non-CRC Non-CRC CONT-B2 Negative Non-CRC Non-CRC CONT-B27 Negative Non-CRC Non-CRC CONT-B3 Negative Non-CRC Non-CRC CONT-B28 Negative Non-CRC Non-CRC CONT-B4 Negative Non-CRC Non-CRC CONT-B29 Negative Non-CRC Non-CRC CONT-B5 Negative Non-CRC Non-CRC CONT-B30 Negative Non-CRC Non-CRC CONT-B6 Negative Non-CRC Non-CRC CONT-B31 Negative Non-CRC Non-CRC CONT-B7 Negative Non-CRC Non-CRC CONT-B32 - Non-CRC Non-CRC CONT-B8 Negative Non-CRC Non-CRC CONT-B33 Negative CRC CRC CONT-B9 Negative Non-CRC Non-CRC CONT-B34 Negative CRC Non-CRC CONT-B10 Negative Non-CRC Non-CRC CONT-B35 Negative Non-CRC Non-CRC CONT-B11 Negative Non-CRC Non-CRC CONT-B36 Negative Non-CRC Non-CRC CONT-B12 Negative Non-CRC Non-CRC CONT-B37 Negative Non-CRC Non-CRC CONT-B13 Negative Non-CRC Non-CRC CONT-B38 Negative Non-CRC Non-CRC CONT-B14 Negative Non-CRC Non-CRC CONT-B39 Negative Non-CRC Non-CRC CONT-B15 Negative Non-CRC Non-CRC CONT-B40 Negative Non-CRC Non-CRC CONT-B16 Negative Non-CRC Non-CRC CONT-B41 Negative Non-CRC Non-CRC CONT-B17 Negative Non-CRC Non-CRC CONT-B42 Negative Non-CRC Non-CRC CONT-B18 Negative Non-CRC Non-CRC CONT-B43 Negative Non-CRC Non-CRC CONT-B19 Negative Non-CRC Non-CRC CONT-B44 Negative Non-CRC Non-CRC CONT-B20 Negative Non-CRC Non-CRC CONT-B45 Negative Non-CRC Non-CRC CONT-B21 Negative Non-CRC Non-CRC CONT-B46 Negative Non-CRC Non-CRC CONT-B22 Negative Non-CRC Non-CRC CONT-B47 Negative Non-CRC Non-CRC CONT-B23 Negative CRC Non-CRC CONT-B48 Negative Non-CRC Non-CRC CONT-B24 Negative Non-CRC Non-CRC CONT-B49 Negative Non-CRC Non-CRC CONT-B25 Negative Non-CRC Non-CRC CONT-B50 Negative Non-CRC Non-CRC CRC FOBT Prediction CRC FOBT Prediction CRC LOME 1-104 CRC LOME 1-104 & CRC LOME 2-23 CRC LOME 1-104 CRC LOME 1-104 & CRC LOME 2-23 CRC-A1 Negative CRC CRC CRC-A140 - CRC CRC CRC-A2 - CRC CRC CRC-A141 Positive CRC CRC CRC-A3 Negative CRC CRC CRC-A142 Negative CRC CRC CRC-A4 Negative CRC CRC CRC-A143 - CRC CRC CRC-A5 Positive CRC CRC CRC-A144 - CRC CRC CRC-A6 Negative CRC CRC CRC-A145 - CRC CRC CRC-A7 - CRC CRC CRC-A146 Positive CRC CRC CRC-A8 Negative CRC CRC CRC-A147 Negative CRC CRC CRC-A9 - CRC CRC CRC-A148 - CRC CRC CRC-A10 Negative CRC CRC CRC-A149 - CRC CRC CRC-A11 Positive CRC CRC CRC-A150 Positive CRC CRC CRC-A12 Positive CRC CRC CRC-A151 Negative CRC CRC CRC-A13 Negative CRC CRC CRC-A152 - CRC CRC CRC-A14 Positive CRC CRC CRC-A153 Negative CRC CRC CRC-A15 Negative CRC CRC CRC-A154 Negative CRC CRC CRC-A16 Negative CRC CRC CRC-A155 - CRC CRC CRC-A17 Positive CRC CRC CRC-A156 Positive CRC CRC CRC-A18 Negative CRC CRC CRC-A157 - CRC CRC CRC-A19 Positive CRC CRC CRC-A158 Negative CRC CRC CRC-A20 Negative CRC CRC CRC-A159 Negative CRC CRC CRC-A21 Negative CRC CRC CRC-A160 Negative CRC CRC CRC-A22 Negative CRC CRC CRC-A161 Positive CRC Non-CRC CRC-A23 Positive CRC CRC CRC-A162 Positive CRC CRC CRC-A24 Negative CRC CRC CRC-A163 - CRC CRC CRC-A25 Positive CRC CRC CRC-A164 Positive CRC CRC CRC-A26 - CRC CRC CRC-A165 Negative CRC CRC CRC-A27 Negative CRC CRC CRC-A166 Negative CRC CRC CRC-A28 Negative CRC CRC CRC-A167 Negative CRC CRC CRC-A29 - Non-CRC Non-CRC CRC-A168 - CRC CRC CRC-A30 - CRC CRC CRC-A169 Positive CRC CRC CRC-A31 Positive CRC CRC CRC-A170 - CRC CRC CRC-A32 - CRC CRC CRC-A171 Negative CRC CRC CRC-A33 Negative CRC CRC CRC-A172 Positive Non-CRC Non-CRC CRC-A34 - CRC CRC CRC-A173 Positive Non-CRC Non-CRC CRC-A35 Negative CRC CRC CRC-A174 - CRC CRC CRC-A36 Negative CRC CRC CRC-A175 - CRC CRC CRC-A37 Positive CRC CRC CRC-A176 Negative CRC CRC CRC-A38 Negative CRC CRC CRC-A177 Negative CRC CRC CRC-A39 Negative CRC CRC CRC-A178 - CRC CRC CRC-A40 - CRC CRC CRC-A179 Positive CRC CRC CRC-A41 - CRC CRC CRC-A180 - CRC CRC CRC-A42 Positive CRC CRC CRC-A181 Negative CRC CRC CRC-A43 Positive CRC CRC CRC-A182 - CRC CRC CRC-A44 Positive CRC CRC CRC-A183 Negative CRC CRC CRC-A45 Positive CRC CRC CRC-A184 Positive CRC CRC CRC-A46 Positive CRC CRC CRC-A185 Positive CRC CRC CRC-A47 Positive CRC CRC CRC-A186 - CRC CRC CRC-A48 Negative CRC CRC CRC-A187 Positive CRC CRC CRC-A49 Positive CRC CRC CRC-A188 - CRC CRC CRC-A50 Positive CRC CRC CRC-A189 Negative CRC CRC CRC-A51 Negative CRC CRC CRC-A190 Negative CRC CRC CRC-A52 - CRC CRC CRC-A191 Positive CRC CRC CRC-A53 Positive CRC CRC CRC-A192 Positive CRC CRC CRC-A54 Negative CRC CRC CRC-A193 Negative CRC CRC CRC-A55 Positive CRC CRC CRC-A194 Negative CRC CRC CRC-A56 Positive CRC CRC CRC-A195 Negative CRC CRC CRC-A57 Positive CRC CRC CRC-A196 Positive CRC Non-CRC CRC-A58 Positive CRC CRC CRC-A197 - CRC CRC CRC-A59 Negative CRC CRC CRC-A198 Positive CRC CRC CRC-A60 - CRC CRC CRC-A199 - CRC CRC CRC-A61 Positive CRC CRC CRC-A200 - CRC CRC CRC-A62 - CRC CRC CRC-A201 - CRC CRC CRC-A63 Negative CRC CRC CRC-A202 Negative CRC CRC CRC-A64 Negative CRC CRC CRC-A203 - CRC CRC CRC-A65 Negative CRC CRC CRC-A204 Negative CRC CRC CRC-A66 Positive CRC CRC CRC-A205 Negative CRC CRC CRC-A67 Negative CRC CRC CRC-A206 Negative CRC CRC CRC-A68 Negative CRC CRC CRC-A207 Positive CRC CRC CRC-A69 Positive CRC CRC CRC-A208 Negative CRC CRC CRC-A70 Positive CRC CRC CRC-A209 - CRC CRC CRC-A71 Positive CRC CRC CRC-A210 Negative CRC CRC CRC-A72 Positive Non-CRC Non-CRC CRC-A211 - CRC CRC CRC-A73 - CRC CRC CRC-A212 - CRC CRC CRC-A74 - CRC CRC CRC-A213 - CRC CRC CRC-A75 - Non-CRC Non-CRC CRC-A214 - CRC CRC CRC-A76 Positive CRC CRC CRC-A215 Negative CRC CRC CRC-A77 - CRC CRC CRC-A216 Positive CRC CRC CRC-A78 Positive CRC CRC CRC-A217 - CRC CRC CRC-A79 Positive Non-CRC Non-CRC CRC-A218 - CRC CRC CRC-A80 Positive CRC CRC CRC-A219 Negative CRC CRC CRC-A81 Positive Non-CRC Non-CRC CRC-A220 - CRC CRC CRC-A82 - CRC CRC CRC-A221 - CRC CRC CRC-A83 Positive CRC CRC CRC-A222 Negative CRC CRC CRC-A84 Positive CRC CRC CRC-A223 - CRC CRC CRC-A85 Positive CRC CRC CRC-A224 Negative CRC CRC CRC-A86 Positive CRC CRC CRC-A225 - CRC CRC CRC-A87 - CRC CRC CRC-A226 - CRC CRC CRC-A88 Negative CRC CRC CRC-A227 - CRC CRC CRC-A89 - CRC CRC CRC-A228 - CRC CRC CRC-A90 Positive CRC CRC CRC-A229 Positive CRC CRC CRC-A91 Negative CRC CRC CRC-A230 Negative CRC CRC CRC-A92 Positive CRC CRC CRC-A231 Negative CRC CRC CRC-A93 Negative CRC CRC CRC-A232 - CRC CRC CRC-A94 - CRC CRC CRC-A233 Positive CRC CRC CRC-A95 - CRC CRC CRC-A234 Positive CRC CRC CRC-A96 Positive CRC CRC CRC-A235 - CRC CRC CRC-A97 Positive CRC CRC CRC-A236 Positive CRC CRC CRC-A98 Positive CRC CRC CRC-A237 Positive CRC CRC CRC-A99 Positive CRC CRC CRC-A238 Negative CRC CRC CRC-A100 Negative CRC CRC CRC-A239 Negative Non-CRC Non-CRC CRC-A101 - Non-CRC Non-CRC CRC-A240 Negative CRC CRC CRC-A102 Negative CRC CRC CRC-A241 Negative CRC CRC CRC-A103 Positive CRC CRC CRC-A242 - CRC CRC CRC-A104 Positive CRC CRC CRC-A243 - CRC CRC CRC-A105 Positive CRC CRC CRC-A244 Positive CRC Non-CRC CRC-A106 Negative CRC CRC CRC-A245 - CRC CRC CRC-A107 Positive CRC CRC CRC-A246 - CRC CRC CRC-A108 Positive CRC CRC CRC-A247 Negative CRC CRC CRC-A109 Positive CRC CRC CRC-A248 - Non-CRC Non-CRC CRC-A110 Positive CRC CRC CRC-A249 - CRC CRC CRC-A111 Positive CRC CRC CRC-A250 Negative CRC CRC CRC-A112 - CRC CRC CRC-A251 Positive CRC CRC CRC-A113 Positive CRC CRC CRC-A252 Negative CRC CRC CRC-A114 - CRC CRC CRC-A253 Positive CRC CRC CRC-A115 Positive CRC CRC CRC-A254 Positive CRC CRC CRC-A116 - CRC CRC CRC-A255 Negative Non-CRC Non-CRC CRC-A117 Positive CRC CRC CRC-A256 Negative Non-CRC Non-CRC CRC-A118 Positive CRC CRC CRC-A257 Negative CRC CRC CRC-A119 Negative CRC CRC CRC-A258 Positive CRC CRC CRC-A120 - CRC CRC CRC-A259 Positive CRC CRC CRC-A121 - CRC CRC CRC-A260 - CRC CRC CRC-A122 Positive CRC Non-CRC CRC-A261 Negative Non-CRC Non-CRC CRC-A123 Positive CRC CRC CRC-A262 Negative Non-CRC Non-CRC CRC-A124 Positive CRC CRC CRC-A263 Negative Non-CRC Non-CRC CRC-A125 Positive CRC CRC CRC-A264 Positive CRC CRC CRC-A126 Negative CRC CRC CRC-A265 Positive CRC CRC CRC-A127 Positive Non-CRC Non-CRC CRC-A266 Positive CRC CRC CRC-A128 Positive CRC CRC CRC-A267 - CRC CRC CRC-A129 - CRC CRC CRC-A268 Positive CRC CRC CRC-A130 - CRC CRC CRC-A269 - CRC CRC CRC-A131 Positive CRC CRC CRC-A270 - CRC CRC CRC-A132 - CRC CRC CRC-A271 - CRC CRC CRC-A133 Positive CRC CRC CRC-A272 Positive Non-CRC Non-CRC CRC-A134 Positive CRC CRC CRC-A273 Positive Non-CRC Non-CRC CRC-A135 Positive CRC CRC CRC-A274 Negative Non-CRC Non-CRC CRC-A136 Negative CRC CRC CRC-A275 - CRC CRC CRC-A137 Negative CRC CRC CRC-A276 Negative CRC CRC CRC-A138 Positive CRC CRC CRC-A277 Positive CRC CRC CRC-A139 Negative CRC CRC CRC-B1 - CRC CRC CRC-B73 Negative CRC CRC CRC-B2 Positive CRC CRC CRC-B74 - CRC CRC CRC-B3 Positive CRC CRC CRC-B75 Negative CRC CRC CRC-B4 Positive CRC CRC CRC-B76 Negative CRC CRC CRC-B5 Negative Non-CRC Non-CRC CRC-B77 - CRC CRC CRC-B6 - CRC CRC CRC-B78 - CRC CRC CRC-B7 Negative CRC Non-CRC CRC-B79 Positive CRC CRC CRC-B8 Positive CRC CRC CRC-B80 Negative CRC CRC CRC-B9 Positive CRC CRC CRC-B81 Negative CRC CRC CRC-B10 Negative CRC Non-CRC CRC-B82 Negative CRC CRC CRC-B11 Positive CRC CRC CRC-B83 Positive CRC CRC CRC-B12 Negative CRC CRC CRC-B84 - CRC CRC CRC-B13 - CRC CRC CRC-B85 Positive CRC CRC CRC-B14 Negative CRC CRC CRC-B86 Positive CRC CRC CRC-B15 - CRC CRC CRC-B87 Negative CRC CRC CRC-B16 Negative CRC CRC CRC-B88 Negative CRC CRC CRC-B17 Negative CRC CRC CRC-B89 Positive CRC CRC CRC-B18 Negative CRC CRC CRC-B90 Positive CRC CRC CRC-B19 - CRC Non-CRC CRC-B91 Negative CRC CRC CRC-B20 Positive CRC CRC CRC-B92 Positive CRC CRC CRC-B21 Negative CRC CRC CRC-B93 - CRC CRC CRC-B22 Negative CRC CRC CRC-B94 - CRC CRC CRC-B23 Positive CRC CRC CRC-B95 Negative CRC Non-CRC CRC-B24 Negative CRC CRC CRC-B96 - CRC Non-CRC CRC-B25 - CRC CRC CRC-B97 - CRC CRC CRC-B26 - CRC CRC CRC-B98 Positive CRC CRC CRC-B27 Negative CRC CRC CRC-B99 Positive CRC CRC CRC-B28 Positive CRC CRC CRC-B100 - CRC CRC CRC-B29 - CRC CRC CRC-B101 - CRC CRC CRC-B30 Positive CRC CRC CRC-B102 Positive CRC CRC CRC-B31 Positive CRC CRC CRC-B103 Negative CRC CRC CRC-B32 Negative CRC CRC CRC-B104 Negative CRC CRC CRC-B33 - CRC CRC CRC-B105 Positive CRC CRC CRC-B34 Positive CRC CRC CRC-B106 Positive CRC CRC CRC-B35 Positive CRC CRC CRC-B107 - CRC CRC CRC-B36 - CRC CRC CRC-B108 Positive CRC CRC CRC-B37 Negative CRC CRC CRC-B109 - CRC CRC CRC-B38 Positive Non-CRC Non-CRC CRC-B110 - CRC CRC CRC-B39 Negative CRC CRC CRC-B111 Negative CRC CRC CRC-B40 Positive CRC CRC CRC-B112 - CRC CRC CRC-B41 - CRC Non-CRC CRC-B113 Negative CRC CRC CRC-B42 Positive CRC CRC CRC-B114 Negative CRC CRC CRC-B43 - CRC CRC CRC-B115 Positive CRC CRC CRC-B44 - CRC CRC CRC-B116 Negative CRC CRC CRC-B45 Negative CRC CRC CRC-B117 Positive CRC CRC CRC-B46 Positive CRC CRC CRC-B118 - CRC CRC CRC-B47 Positive CRC CRC CRC-B119 Positive CRC CRC CRC-B48 Negative CRC CRC CRC-B120 Negative CRC CRC CRC-B49 Negative CRC CRC CRC-B121 - CRC CRC CRC-B50 - CRC CRC CRC-B122 Positive CRC CRC CRC-B51 Negative CRC CRC CRC-B123 - CRC CRC CRC-B52 - CRC CRC CRC-B124 Negative CRC CRC CRC-B53 - CRC CRC CRC-B125 Positive CRC CRC CRC-B54 Positive CRC CRC CRC-B126 - CRC CRC CRC-B55 - CRC CRC CRC-B127 - CRC CRC CRC-B56 - CRC CRC CRC-B128 Negative CRC CRC CRC-B57 Positive CRC CRC CRC-B129 Negative CRC CRC CRC-B58 Positive CRC CRC CRC-B130 - CRC CRC CRC-B59 - CRC CRC CRC-B131 - CRC CRC CRC-B60 Negative CRC CRC CRC-B132 Negative CRC CRC CRC-B61 Positive CRC CRC CRC-B133 - CRC CRC CRC-B62 Positive CRC CRC CRC-B134 Positive CRC CRC CRC-B63 Negative CRC CRC CRC-B135 - CRC CRC CRC-B64 Negative CRC CRC CRC-B136 - CRC CRC CRC-B65 Negative CRC CRC CRC-B137 Positive CRC CRC CRC-B66 Positive CRC CRC CRC-B138 Positive CRC CRC CRC-B67 Negative CRC CRC CRC-B139 Positive CRC CRC CRC-B68 Negative CRC CRC CRC-B140 - CRC CRC CRC-B69 - CRC CRC CRC-B141 Negative CRC CRC CRC-B70 Positive CRC CRC CRC-B142 Positive CRC CRC CRC-B71 - CRC CRC CRC-B143 - CRC CRC CRC-B72 Negative CRC CRC

In order to confirm the reproducibility of the discrimination results according to the present invention, the same procedure was applied to 13 normal controls, 35 CRC patients, 7 BRC patients, 14 gastric cancer patients, 7 OVC patients and 5 TA patients The results are shown in Table 32. When the result of the discrimination was reversed, it was concentrated in the TA patient group. Tis in the TA was classified as cancer or not classified, and the clinical result was confusing, and the discrimination result according to the present invention showed confusing results. It can be confirmed that the reproducibility is 90% or more, which demonstrates the excellent discrimination performance of the present invention.

Prediction after
1st 5 times repeated
Low mass ion measurements Prediction after
2nd 5 times repeated
Low mass ion measurements Prediction after
1st 5 times repeated
Low mass ion measurements Prediction after
2nd 5 times repeated
Low mass ion measurements CRC LOME 1-104
& CRC LOME 2-23 CRC LOME 1-104
& CRC LOME 2-23 CRC LOME 1-104
& CRC LOME 2-23 CRC LOME 1-104
& CRC LOME 2-23 CONT-B1 Non-CRC Non-CRC CRC-B1 CRC CRC CONT-B5 Non-CRC Non-CRC CRC-B2 CRC CRC CONT-B9 Non-CRC Non-CRC CRC-B3 CRC CRC CONT-B13 Non-CRC Non-CRC CRC-B4 CRC CRC CONT-B17 Non-CRC Non-CRC CRC-B5 Non-CRC Non-CRC CONT-B21 Non-CRC Non-CRC CRC-B7 Non-CRC CRC CONT-B26 Non-CRC Non-CRC CRC-B8 CRC CRC CONT-B29 Non-CRC Non-CRC CRC-B9 CRC CRC CONT-B33 CRC Non-CRC CRC-B10 Non-CRC CRC CONT-B36 Non-CRC Non-CRC CRC-B11 CRC CRC CONT-B41 Non-CRC Non-CRC CRC-B12 CRC CRC CONT-B47 Non-CRC Non-CRC CRC-B13 CRC CRC CONT-B49 Non-CRC Non-CRC CRC-B15 CRC CRC BRC-B1 Non-CRC Non-CRC CRC-B16 CRC CRC BRC-B5 Non-CRC Non-CRC CRC-B18 CRC CRC BRC-B9 Non-CRC Non-CRC CRC-B19 Non-CRC CRC BRC-B13 Non-CRC Non-CRC CRC-B20 CRC CRC BRC-B17 Non-CRC Non-CRC CRC-B22 CRC CRC BRC-B21 Non-CRC Non-CRC CRC-B24 CRC CRC BRC-B25 Non-CRC Non-CRC CRC-B25 CRC CRC GC-B1 Non-CRC Non-CRC CRC-B26 CRC CRC GC-B5 Non-CRC Non-CRC CRC-B28 CRC CRC GC-B9 Non-CRC Non-CRC CRC-B29 CRC CRC GC-B13 Non-CRC Non-CRC CRC-B32 CRC CRC GC-B17 Non-CRC Non-CRC CRC-B33 CRC CRC GC-B21 Non-CRC Non-CRC CRC-B34 CRC CRC GC-B25 CRC Non-CRC CRC-B35 CRC Non-CRC GC-B29 Non-CRC Non-CRC CRC-B36 CRC CRC GC-B33 Non-CRC Non-CRC CRC-B37 CRC CRC GC-B37 Non-CRC Non-CRC CRC-B38 Non-CRC Non-CRC GC-B41 Non-CRC Non-CRC CRC-B39 CRC CRC GC-B45 Non-CRC Non-CRC CRC-B40 CRC CRC GC-B49 Non-CRC Non-CRC CRC-B41 Non-CRC CRC GC-B53 Non-CRC Non-CRC CRC-B42 CRC CRC OVC-B1 Non-CRC Non-CRC CRC-B43 CRC CRC OVC-B5 Non-CRC Non-CRC TA-B1 CRC Non-CRC OVC-B9 CRC CRC TA-B5 CRC Non-CRC OVC-B13 Non-CRC Non-CRC TA-B8 CRC Non-CRC OVC-B17 Non-CRC Non-CRC TA-B14 CRC Non-CRC OVC-B21 Non-CRC Non-CRC TA-B18 Non-CRC Non-CRC OVC-B25 Non-CRC Non-CRC Reproducibility 90.79% (86.42%, when TA was included)

FIGS. 20A and 20B show results of identifying 1465.6184 m / z and 2450.9701 m / z, respectively, among low mass ions for the first kind CRC diagnosis. Both of the two low mass ions show the shape of the peak mass of the same material with a mass value difference of about 1 m / z depending on the number of constituent isotopes. This is a mass peak pattern specific to a protein or peptide appearing in a mass spectrometer.

20A and 20B show the mass spectra of each of the above two low mass ions. The spectrum indicated by the red line represents the peak intensity in the serum extract of the CRC patients and the spectrum indicated by the blue line is obtained from the normal control group. 1465.6184 m / z showed higher peak intensity in CRC patients whereas 2450.9701 m / z showed lower peak intensity in CRC patients. 20A and 20B show MS / MS analysis spectra of the above two low mass ions, and the tables in FIGS. 20A and 20B and the lower left graph show that 1465.6184 m / z and 2450.9701 m / The analysis shows that each of them was identified as a fibrinogen alpha chain and a trans-retinin protein.

Quantitative measurement of the peak intensity of low mass ion 1465.6184 m / z corresponding to the fibrinogen alpha chain in the CRC patients showed that the level of fibrinogen in the blood of the CRC patients was higher than that of the normal ones, (See Table 33).

Number Plasma fibrinogen level (mg / dL), mean ± standard deviation p-value Healthy control 37 274.51 + - 93.22 <0.001 Colorectal adenoma 29 279.59 ± 58.03 1,000 Stage I CRC 148 298.93 + - 69.40 0.685 Stage II CRC 340 351.14 + - 96.65 <0.001 Stage III CRC 507 345.19 ± 95.25 <0.001 Stage IV CRC 57 365.33 + - 91.37 <0.001

On the other hand, quantitative measurement of the peak intensity of low mass ion 2450.9701 m / z corresponding to the trans-retin, corresponding to the low qualitative result in the CRC patient group, showed that the blood trans-retinol level in the CRC patient (See Table 34). The mean ± standard deviation was 160.39 ± 62.41 ng / mL in CRC patients and 171.19 ± 30.86 ng / mL in normal controls.

Level of Transthyretin
(ng / mL) Level of Transthyretin
(ng / mL) CRC-A134 194.4053871 CRC-B25 163.7544561 CRC-A135 160.3388216 CRC-B26 103.763308 CRC-A137 45.31734887 CRC-B27 272.5678515 CRC-A139 154.2433081 CRC-B28 146.0108407 CRC-A140 201.5848401 CRC-B29 157.6719758 CRC-A141 181.6259276 CRC-B30 34.24514756 CRC-A142 181.8730938 CRC-B32 209.2664543 CRC-A143 209.4562651 CRC-B33 14.23188743 CRC-A144 204.9015183 CRC-B34 170.7668514 CRC-A145 204.8086556 CRC-B35 240.8403292 CRC-A146 151.7466189 CRC-B46 314.1906603 CRC-A147 155.5422654 CRC-B47 24.56925637 CRC-A148 241.4415416 CRC-B48 201.2652628 CRC-A149 232.1575272 CONT-B2 131.7631966 CRC-A150 178.8075456 CONT-B3 185.5791386 CRC-A151 150.5475887 CONT-B4 118.2875787 CRC-A152 150.3770739 CONT-B5 134.204582 CRC-A153 140.3009368 CONT-B6 122.8785828 CRC-A154 146.5080959 CONT-B7 163.1616804 CRC-A155 20.11742957 CONT-B8 155.4717726 CRC-A156 148.9523525 CONT-B9 234.9631822 CRC-B2 107.8883621 CONT-B10 162.5710506 CRC-B4 176.2724527 CONT-B11 162.7922858 CRC-B5 39.91544623 CONT-B12 159.0358497 CRC-B6 156.5325266 CONT-B14 251.1537438 CRC-B7 199.4036008 CONT-B15 160.8484111 CRC-B10 145.7793662 CONT-B16 157.2437136 CRC-B11 181.4202125 CONT-B18 187.9070482 CRC-B12 158.8557207 CONT-B19 158.4960747 CRC-B13 176.0328979 CONT-B20 184.9911989 CRC-B14 209.6462481 CONT-B22 177.2741119 CRC-B16 142.1897289 CONT-B23 151.5403947 CRC-B18 224.0415149 CONT-B24 203.7437549 CRC-B19 212.08369 CONT-B27 179.7828571 CRC-B20 179.8236103 CONT-B32 199.8560708 CRC-B21 154.8387737 CONT-B34 185.7895696 CRC-B22 156.8521618 CONT-B35 163.8658411 CRC-B23 89.36660383 CONT-B48 158.9637736 CRC-B24 120.9443976 CONT-B50 198.7268159

According to one embodiment of the present invention, a low mass ion mass spectrum of serum was analyzed, and CRC patients and non-patients could be discriminated with high discrimination ability.

The present invention can be easily expanded to form a formula for discriminating between a specific cancer patient group and a normal control group other than colorectal cancer through a process similar to that of the present embodiment, and also constituting a discrimination formula for discriminating different cancer patient groups. It will also be readily apparent to those of skill in the art that it can be readily extended to diagnose cancer as well as any other disease.

1000: low mass ion detection unit 2000: colon cancer diagnosis unit
2100: first alignment means 2200: first discrimination score calculation means
2210: Normalization module 2220: Scaling module
2230: discrimination score calculation module 2300: factor accumulation value calculation means
2310: First training set selection means 2400: Colon cancer diagnosis ion selection means
2410: Candidate ion set selection means 2411: First low mass ion selection module
2412: Candidate ion set preliminary selection module 2413: Sensitivity and specificity calculation module
2414: Candidate ion set final selection module 2420: Final ion set selection means
2421: Ionizing module 2422: Biomarker group pre-selection module
2423: Biomarker group re-selection module 2424: Biomarker group final selection module
2425: Biomarker group addition module 2426: Low mass ion selective module for cancer diagnosis
2500: second aligning means 2600: second discrimination score calculating means
2610: Normalization module 2620: Scaling module
2630: discrimination score calculation module 2700: colon cancer determination means
3000:

Claims (37)

A low mass ion detector for detecting mass spectra of low-mass ions from biological samples of a plurality of colorectal cancer patients and non-patient cases;
A colon cancer diagnosis unit for comparing and analyzing the mass spectral patterns of the low-mass ions to determine colon cancer diagnosis information; And
And a display unit for converting the colorectal cancer diagnosis information determined by the colorectal cancer diagnosis unit into an outputable form and displaying the colorimetric cancer diagnosis information. The method according to claim 1,
Wherein the low mass ion detector comprises:
And a peak intensity of a low-mass ion is detected from the biological sample to extract a mass spectrum of the low-mass ion. 3. The method of claim 2,
Wherein the low mass ion detector comprises:
And a mass analyzer. The method according to claim 1,
Wherein the colon cancer diagnosis unit comprises:
First alignment means for aligning low mass ion mass spectra of said colorectal cancer patients and non-patient cases which are training candidate sets;
A first discrimination score calculating means for calculating a discrimination score by performing a biostatistical analysis on the sorted mass spectrum;
A factor load calculation means for calculating a sensitivity and a specificity according to the discrimination score and calculating a factor loading by selecting a first training set based on the sensitivity and the specificity,
An ion selection means for diagnosing colon cancer which selects low mass ions for diagnosis of colon cancer based on the discrimination performance among candidate low mass ions satisfying the candidate condition;
Second alignment means for aligning the low mass ion mass spectrum of the identified biological sample with the first training set;
A second discrimination score calculation means for calculating a discrimination score from the peak intensity of the low-mass ion of the discrimination object and the factor load value; And
And colorectal cancer determination means for determining the discrimination target as positive or negative for colorectal cancer based on the discrimination score,
The colon cancer diagnostic ion selecting means comprises:
The plurality of colorectal cancer patients and non-patient cases are classified into first type discrimination cases each comprising a plurality of colon cancer patient cases and a plurality of normal case cases, And second type discrimination cases composed of cancer patient cases of the second kind,
The low-mass ions for diagnosing colorectal cancer are subjected to the first type discrimination cases and the second type discrimination cases, respectively, and the low-mass ions for diagnosis of the first type colorectal cancer and the low- And a low-mass ion for diagnosing second-type colorectal cancer for the second-type discriminating case. 5. The method of claim 4,
The mass values of the low mass ions for diagnosis of the first type colorectal cancer are 18.0260, 22.9797, 74.0948, 76.0763, 102.0916, 105.1078, 106.0899, 107.0477, 118.0822, 123.0395, 137.0423, 137.0729, 147.0573, 147.1058, 169.0653, 181.0656, 190.0849, 191.0848, 369.3302, 377.0570, 369.3302, 357.2781, 367.2669, 367.2784, 366.2310, 368.2551, 369.3302, 377.0570, 290.036, 379.4765, 383.0529, 384.1745, 388.2688, 401.0531, 423.0313, 428.1878, 454.2090, 465.3014, 466.1923, 468.1851, 469.2831, 477.1721, 478.1678, 480.1715, 482.3220, 483.3258, 496.8683, 497.7636, 503.8719, 508.3407, 510.3265, 512.3119, 513.3177, 518.2931, 518.8555, 519.2967, 519.8598, 525.3449, 534.2739, 537.2800, 538.3306, 540.2629, 540.8144, 542.8457, 544.8692, 548.2856, 566.8375, 581.1957, 582.1888, 583.2242, 656.0270, 667.3291, 709.3519, 710.3581, 711.3617, 712.3683, 713.3798, 991.6196, 992 Characterized in that the mass value is at least one mass value selected from the group consisting of m / z of 6209, 1016.6113, 1020.4817, 1206.5305, 1207.5571, 1465.6184, 1466.6096, 1467.5969, 2450.9701, 2451.9662 and 2452.9546 m / z A colon cancer diagnostic device. 5. The method of claim 4,
The cancer patient cases other than the plurality of colorectal cancer among the second type discrimination cases include at least one cancer patient case, non-Hodgkin lymphoma (NHL) patient case, and gastric cancer patient case &Lt; / RTI &gt;
The mass values of the second kind of colorectal cancer diagnostic low mass ions are 60.0476, 138.0540, 172.6653, 173.1158, 179.1451, 191.1277, 279.0855, 280.0895, 280.2642, 281.1440, 296.2574, 312.3248, 332.3224, 333.3324, 369.3406, 465.3161, 486.6356, 488.6882, 544.8908, 551.3287, 566.8737, 707.3475 and 733.3569 m / z (with an error range of +/- 0.1 m / z). 5. The method of claim 4,
Wherein the discrimination score includes a first type discrimination score by the first kind low mass ions and a second type discrimination score by the second kind low mass ions,
When the first type discrimination score is greater than CS ₁₁ and the second type discrimination score is greater than CS ₂₁ , the discrimination object is judged to be colon cancer positive,
If the first type discrimination score is smaller than CS ₁₂ or the second type discrimination score is smaller than CS ₂₂ , the discrimination object is judged as negative colon cancer. 8. The method of claim 7,
Wherein the CS ₁₁ , CS ₁₂ , CS _21, and CS ₂₂ are zero. 5. The method of claim 4,
Wherein the low-mass ion group that distinguishes colon cancer patients and normal persons obtained using the ion selection means for colon cancer diagnosis includes fibrinogen or fibrinogen alpha chain. 5. The method of claim 4,
Wherein the low-mass ion group that distinguishes between the colon cancer patient and the normal person includes transthyretin. 5. The method of claim 4,
Wherein the first discrimination score calculating means comprises:
A normalization module for normalizing the peak intensities of the low mass ion mass spectra of the training candidate set;
A scaling module for scaling the normalized peak intensities; And
And a discrimination score calculation module for calculating the discrimination score by performing the biostatistical analysis on the scaled peak intensities. 12. The method of claim 11,
The scaling module comprising:
And performing Pareto scaling. 12. The method of claim 11,
Wherein the discrimination score calculation module comprises:
Wherein the biostatistical analysis is performed using linear discriminant analysis based on principal component analysis. 14. The method of claim 13,
Wherein the discrimination score calculation module comprises:
Wherein the discrimination score is calculated using both the factor load value obtained by the principal component analysis based linear discriminant analysis and the scaled peak intensity together. 5. The method of claim 4,
The factor load value calculation means comprising:
Performing a biostatistical analysis on the sorted mass spectrum and selecting training cases that satisfy training conditions among the colorectal cancer patients and non-patient cases based on the biostatistical analysis results as a first training set 1 training set selection means, and calculates the parameter load value from the first training set. 16. The method of claim 15,
Wherein the first training set selection means comprises:
Wherein colon cancer patients and non-cancer cases when the sensitivity according to the result of the biostatistical analysis is equal to or greater than the threshold value CN ₁ and the specificity is equal to or greater than the threshold value CN ₂ are set as the first training set. Device. 17. The method of claim 16,
Wherein the threshold values CN ₁ and CN ₂ are 1. 5. The method of claim 4,
And the second discrimination score calculating means comprises:
A normalization module for normalizing the peak intensities of the low mass ion mass spectrum of the object to be discriminated;
A scaling module for scaling the normalized peak intensities; And
And a discrimination score calculation module for calculating the discrimination score through the scaled peak intensity and the factor load value. 19. The method of claim 18,
The scaling module comprising:
And performing Pareto scaling. 19. The method of claim 18,
Wherein the discrimination score calculation module comprises:
Wherein the discrimination score is calculated through the scaled peak intensity of the low mass ions for colorectal cancer diagnosis and the factor load value. 5. The method of claim 4,
The colorectal cancer determining means comprises:
Determining whether the discriminant is positive or negative for colorectal cancer according to the discrimination score,
If the discrimination score is greater than the reference value CS, colon cancer information of the discrimination object is judged to be positive, and if the discrimination score is lower than the reference value CS, colon cancer information of the discrimination subject is judged by voice negatively . 22. The method of claim 21,
Wherein the reference value CS is zero. 5. The method of claim 4,
The colorectal cancer determining means comprises:
Characterized in that the colorectal cancer information of the discrimination object is judged by using an average value of a plurality of the discrimination scores calculated for a plurality of the low mass ion mass spectra obtained by repeatedly measuring the biological sample to be discriminated Diagnostic device. 5. The method of claim 4,
The colon cancer diagnostic ion selecting means comprises:
Candidate ion set selection means for selecting candidate low-mass ions satisfying a candidate condition from the selected first training set as a candidate ion set; And
And final ion collection means for selecting low mass ions for colorectal cancer diagnosis as a final collection of ions based on the discrimination performance of the candidate low mass ions of the selected candidate collection of ions or individual combinations thereof Diagnostic device. 25. The method of claim 24,
The candidate ion set selection means comprises:
Wherein a peak intensity of the low mass ion for each of the training cases and an absolute value of a product of the low mass ion charge accumulation value obtained through the biostatistical analysis is greater than a threshold CT ₁ And a first low mass ion selection module for selecting each of the training cases. 26. The method of claim 25,
Colon cancer diagnosis apparatus as the threshold value CT ₁ is characterized in that 0.1. 26. The method of claim 25,
The candidate ion set selection means comprises:
And a candidate ion collection preliminary selection module for selecting the second low mass ions common to the cases among the first low-mass ions in the cases of CT ₂ percent or more of the entire training cases CT as the candidate ion collection A colon cancer diagnosis device. 28. The method of claim 27,
Colon cancer diagnostic device, characterized in that the threshold value CT ₂ is 50. 28. The method of claim 27,
The candidate ion set selection means comprises:
A sensitivity and specificity calculating module for calculating a discrimination score indicating positive or negative for colorectal cancer in each training case using the second low-mass ions, and calculating sensitivity and specificity according to the discrimination score; And
Candidate ions to the sensitivity of less than or equal to the threshold value CN _3, wherein the specificity is less than the threshold value CN ₄ to change at least one of the CT ₁ and the CT _2, by repeating the process selected for the candidate ion set And a set final selection module. 30. The method of claim 29,
Wherein the thresholds CN ₃ and CN ₄ are 0.9. 25. The method of claim 24,
The criterion of the discrimination performance of the final ion set selection means is:
Wherein the sum of the sensitivity and the specificity among the candidate low mass ions is larger than the reference value or the sum of the sensitivity and the specificity of each combination of the candidate low mass ions is greater than the sum of the sensitivity and the specificity, 1 criteria for colorectal cancer. 25. The method of claim 24,
The criterion of the discrimination performance of the final ion set selection means is:
And a second criterion for selecting a combination of the candidate low-mass ions among the combinations of the candidate low-mass ions as the lowest combination of the comparison-target combinations. 25. The method of claim 24,
The criterion of the discrimination performance of the final ion set selection means is:
The difference between the minimum discrimination score of the true positive case and the maximum discrimination score of the true negative case among the combinations of the candidate low mass ions is the third criterion for selecting the largest combination of the comparison object combinations &Lt; / RTI &
Wherein the discrimination score is calculated through the scaled peak intensity of the candidate low mass ions and the factor load value and represents positive or negative for colorectal cancer. 25. The method of claim 24,
Said final ion set selection means comprising:
Wherein the candidate low-mass ions included in the candidate ion set include high-sensitivity low-mass ions whose sensitivity is higher than the specificity, and high sensitivity that aligns the high-sensitivity low-mass ions in descending order of the sensitivity and the specificity The set {Sns ₁ , Sns ₂ , Sns ₃ ... Sns _I }, a high specificity set {Spc ₁ , Sns _I } comprising the high specificity low mass ions whose specificity is higher than the sensitivity, and sorting the high specific low mass ions in descending order of the sensitivity and the specificity, Spc ₂ , Spc ₃ ... Spc _J };
The high-sensitivity low-mass ions {Sns ₁ , Sns ₂ , Sns ₃ ... Sns _L } and the high-order L low-mass ions {Spc ₁ , Spc ₂ , Spc ₃ ... Of the candidate combination consisting of Spc _L} of two or more of the low mass ions the first reference and the second reference and the third the combination selected by the criterion for the discrimination performance by at least one of the reference biomarker A biomarker group pre-selection module to be selected as a group;
Wherein said biomarker group comprises at least one of the above-mentioned high-sensitivity low-mass ions at the next higher-order M in said hypertonic sensitivity set and at least one of said high-specific low- A re-selection of a biomarker group reassigning the combination selected by the criterion of the discrimination performance by at least one of the first criterion, the second criterion and the third criterion among the candidate combinations added to the group to the biomarker group module; And
And a final biomarker group final selection module for selecting the biomarker group by repeating the re-selection process until there is no next-low mass ion in the hyper-sensitivity set and the high specificity aggregate. Cancer diagnostic device. 35. The method of claim 34,
Said final ion set selection means comprising:
The remaining candidate ion sets excluding the combination of the low-mass ions selected as the biomarker group obtained in the biomarker group final selection module in the candidate ion set are repeatedly subjected to the selection process of the three biomarker groups, A biomarker group addition selecting module for further selecting the additional biomarker group until the L or less mass ions remain in the hyperhidimetic sensitivity set or the high specificity set; And
Wherein the low mass ions of the combination of the upper CK biomarker groups are selected as the low mass ion for the diagnosis of colorectal cancer based on the accuracy of the true positive and negative negative determinations among the biomarker group and the additional biomarker groups Further comprising a diagnostic low mass ion final selection module. 36. The method of claim 35,
Wherein the L value is 2, the M value is 1, and the CK value is 1, 2, or 3. 25. The method of claim 24,
Wherein said final ion set selection means comprises:
Wherein the low mass ion is selected for a training set in which a second training set independent of the first training set is added to the first training set.

Images