阅读说明：本技术 一种分析靶向测序Panel估算肿瘤突变负荷可行性的方法和装置 (Method and device for analyzing feasibility of target sequencing Panel for estimating tumor mutation load ) 是由 张仕坚 季序我 于 2020-12-31 设计创作，主要内容包括：本发明提供了一种分析靶向测序Panel估算肿瘤突变负荷(TMB)可行性的方法和装置,根据全外显子组(WES)和Panel的基因长度及突变数量,计算WES和Panel TMB；按相同的分组规则,分别根据WES及Panel TMB,将样本分为全外显子组及Panel中的不同高低等级组,并据此计算严重错误分组百分比,分析靶向测序Panel估算TMB的可行性。本发明采用多种方法定量评估WES TMB和Panel估算的TMB的一致性,包括相关性系数和严重错误分组百分比,两者的评估结论一致。本发明的特色之处在于：在按照癌种分类的基础上,进一步按性别或年龄分类,并分析性别或年龄因素对TMB一致性的影响,将临床决策进一步细化,提高决策准确率；对现有Panel进行优化及为构建新Panel提供了方法。(The invention provides a method and a device for analyzing feasibility of target sequencing Panel for estimating tumor mutation load (TMB), wherein WES and Panel TMB are calculated according to gene length and mutation quantity of a Whole Exome (WES) and Panel; according to the same grouping rule, samples are divided into different high-level and low-level groups in the whole exome and Panel according to WES and Panel TMB respectively, the percentage of serious error grouping is calculated according to the groups, and the feasibility of target sequencing Panel for estimating TMB is analyzed. The invention adopts a plurality of methods to quantitatively evaluate the consistency of WES TMB and the TMB estimated by Panel, including the correlation coefficient and the serious error grouping percentage, and the evaluation conclusion of the WES TMB and the TMB estimated by Panel is consistent. The invention is characterized in that: based on classification according to cancer species, further classifying according to gender or age, analyzing the influence of gender or age factors on TMB consistency, further refining clinical decision and improving decision accuracy; the method optimizes the existing Panel and provides a method for constructing a new Panel.)

1. A method for analyzing the feasibility of targeted sequencing Panel for estimating tumor mutation burden, the method comprising:

calculating the tumor mutation load based on the full exome and the tumor mutation load based on the Panel according to the gene length and the mutation number of the full exome and the Panel respectively;

calculating Spearman correlation coefficients for the full exome-based tumor mutation burden and the Panel-based tumor mutation burden;

dividing the sample into a plurality of groups based on the full exome according to the tumor mutation load value of the full exome and a plurality of groups based on Panel according to the tumor mutation load value of Panel according to the same grouping rule, and calculating the percent of serious error grouping according to the grade difference of the sample between the groups based on the full exome and the groups based on Panel;

the feasibility of targeted sequencing Panel to estimate tumor mutation burden was analyzed based on Spearman correlation coefficient and/or percent of severe error groupings.

2. The method of claim 1, wherein the tumor mutational burden is calculated by the formula:

3. the method of claim 1, wherein the grouping rule comprises evenly dividing the samples into no less than 3 levels according to the tumor mutation burden value ordering;

the calculation formula of the serious error grouping percentage Fs is as follows:

Fs＝S/(C+M+S)

wherein S is the number of samples grouped by gross errors, the gross errors being samples that differ by at least two degrees in the full exome-based group and in the Panel-based group;

c is the number of samples correctly grouped, with samples having the same rank in the full exome-based group and in the Panel-based group;

m is the number of samples grouped slightly in error, which is one level difference between samples in the full exome-based group and in the Panel-based group.

4. The method of claim 1, further comprising the step of classifying the sample according to cancer type, optionally gender or age, prior to subjecting the sample to the tumor mutational burden classification.

5. An apparatus for analyzing feasibility of targeted sequencing Panel for estimating tumor mutation burden, the apparatus comprising:

a tumor mutation load calculation module for calculating the tumor mutation load based on the full exome and the tumor mutation load based on the Panel according to the gene length and the mutation number of the full exome and the Panel respectively;

a Spearman correlation coefficient calculation module for calculating a Spearman correlation coefficient for a full exome-based tumor mutation burden and a Panel-based tumor mutation burden;

a serious error grouping percentage calculation module for dividing the sample into a plurality of groups based on the whole exome according to the tumor mutation load value of the whole exome according to the same grouping rule, dividing the sample into a plurality of groups based on Panel according to the tumor mutation load value of Panel, and calculating the serious error grouping percentage according to the grade difference of the sample between the group based on the whole exome and the group based on Panel;

the sample grouping and deviation statistical test module is used for classifying the samples according to cancer species, cancer species and gender or cancer species and age, calling the Spearman correlation coefficient calculation module or the serious error grouping percentage calculation module, calculating the classified Spearman correlation coefficient or the serious error grouping percentage, and then carrying out statistical significance test on the influence of gender or age on consistency;

and the analysis module is used for analyzing the feasibility of estimating the tumor mutation load by the target sequencing Panel according to the Spearman correlation coefficient, the serious error grouping percentage and the statistical significance test result.

6. The apparatus of claim 5, wherein the tumor mutational burden is calculated by the formula:

7. the apparatus of claim 5, wherein the grouping rule comprises evenly dividing the samples into no less than 3 levels according to the tumor mutation burden value ordering.

8. The apparatus of claim 5, wherein the percentage of fatal error packets Fs is calculated by the formula:

Fs＝S/(C+M+S)

wherein S is the number of samples grouped by gross errors, the gross errors being samples that differ by at least two degrees in the full exome-based group and in the Panel-based group;

c is the number of samples correctly grouped, with samples having the same rank in the full exome-based group and in the Panel-based group;

m is the number of samples grouped slightly in error, which is one level difference between samples in the full exome-based group and in the Panel-based group.

9. A method of screening for a Panel gene to optimize identity, the method comprising:

calculating Spearman correlation coefficients for tumor mutation burden based on a single gene and tumor mutation burden based on a full exome;

screening genes from existing panels for sub-panels based on Spearman correlation coefficients and analyzing said sub-panels for the extent to which the feasibility of estimating tumor mutational burden is further improved;

genes were screened from the whole exome group based on Spearman correlation coefficient to form new panels, and the feasibility of the new panels to estimate tumor mutational burden was analyzed.

10. An apparatus for screening for a Panel gene for identity, said apparatus comprising:

a Spearman correlation coefficient calculation module for calculating a Spearman correlation coefficient for a tumor mutation burden based on a single gene and a tumor mutation burden based on a full exome;

a sub-Panel creation module for screening genes from existing Panel according to a Spearman correlation coefficient to form sub-Panel, and analyzing the sub-Panel to estimate the degree of improvement in the feasibility of tumor mutation burden;

a new Panel creation module for screening genes from the whole exome group based on the Spearman correlation coefficient to form a new Panel, and analyzing the new Panel to estimate the degree of increased feasibility of tumor mutational burden.

Technical Field

The invention belongs to the technical field of bioinformatics, and relates to a method and a device for analyzing feasibility of target sequencing Panel for estimating tumor mutation load.

Background

Tumor Mutation Burden (TMB) has become an important biomarker for developing Tumor immunotherapy based on immune checkpoint inhibitors, and several drugs have been approved for concomitant diagnostic applications based on TMB. Currently, Whole Exome Sequencing data (WES) is mainly used for calculating TMB, and since the WES contains mutation detection information of all human protein coding genes (more than twenty thousand), the TMB can be calculated accurately on the Whole, which is an industry gold standard, but has the problem of high cost. To reduce costs, some products are available on the market that use partial genes (typically hundreds) for TMB estimation, and it is desirable to represent the whole case by sampling, and this set of partial genes is called Panel. Whether the TMB estimated based on patch generated by sampling is consistent with the tms calculated by WES has been published with relevant research efforts. It is widely recognized in the prior art that Panel can estimate TMB to represent WES-calculated TMB as a whole without distinguishing cancer species, and has better representativeness among individual cancer species.

While Panel can represent WES well in all and individual cancer species, it is generally representative of other cancer species, with some even poor (i.e., low correlation coefficients). In practice, Panel is used in a particular cancer species. On the other hand, different panels are available on the market for different cancer species, and the representativeness of WES is different, and even though a specific Panel may show better representativeness, the panels may contain genes which are not suitable for accurately calculating TMB (Tetramethylbenzidine), such as targeted medication guidance, and the like, so that the representativeness cannot reach the optimal degree.

Therefore, we must be evaluated for representativeness of a particular Panel to screen out the optimal Panel and consider applicability under the influence of a particular cancer species and different clinical factors (e.g., gender and age).

Disclosure of Invention

In view of the deficiencies and practical needs of the prior art, the present invention provides, in a first aspect, a method for analyzing the feasibility of targeted sequencing Panel for estimating tumor mutation burden, the method comprising:

calculating the tumor mutation load based on the full exome and the tumor mutation load based on the Panel according to the gene length and the mutation number of the full exome and the Panel respectively;

calculating Spearman correlation coefficients for the full exome-based tumor mutation burden and the Panel-based tumor mutation burden;

dividing the sample into a plurality of groups based on the full exome according to the tumor mutation load value of the full exome and a plurality of groups based on Panel according to the tumor mutation load value of Panel according to the same grouping rule, and calculating the percent of serious error grouping according to the grade difference of the sample between the groups based on the full exome and the groups based on Panel;

the feasibility of targeted sequencing Panel to estimate tumor mutation burden was analyzed based on Spearman correlation coefficient and/or percent of severe error groupings.

In the invention, the consistency of TMB calculated by WES and TMB estimated by Panel is quantitatively evaluated by adopting the percentage of serious error grouping, a result consistent with a correlation coefficient is obtained, which shows that the representativeness of Panel can be evaluated by taking the percentage of serious error grouping as an index, and the index more intuitively reflects the misdiagnosis or missed diagnosis probability of a patient compared with the correlation coefficient.

Preferably, the calculation formula of the tumor mutation load is as follows:

preferably, the grouping rule comprises uniformly classifying the samples into not less than 3 levels according to the tumor mutation load value ordering, and preferably uniformly classifying the samples into high, medium and low 3 levels according to the tumor mutation load value ordering.

Preferably, the calculation formula of the serious error grouping percentage Fs is as follows:

Fs＝S/(C+M+S)

wherein S is the number of samples grouped by gross errors, the gross errors being samples that differ by at least two degrees in the full exome-based group and in the Panel-based group;

c is the number of samples correctly grouped, with samples having the same rank in the full exome-based group and in the Panel-based group;

m is the number of samples grouped slightly in error, which is one level difference between samples in the full exome-based group and in the Panel-based group.

Preferably, before the tumor mutation burden grouping is performed on the samples, the step of classifying the samples according to cancer species, cancer species and sex or cancer species and age is further included.

In the present invention, in view of the poor representativeness of commercial Panel in particular cancer species, it is advantageous to classify samples further according to gender or age to improve the representativeness of Panel in particular population of particular cancer species.

In a second aspect, the present invention provides an apparatus for analyzing feasibility of targeted sequencing Panel for estimating tumor mutation burden, the apparatus comprising:

a tumor mutation load calculation module for calculating the tumor mutation load based on the full exome and the tumor mutation load based on the Panel according to the gene length and the mutation number of the full exome and the Panel respectively;

a Spearman correlation coefficient calculation module for calculating a Spearman correlation coefficient for a full exome-based tumor mutation burden and a Panel-based tumor mutation burden;

a serious error grouping percentage calculation module for dividing the sample into a plurality of groups based on the whole exome according to the tumor mutation load value of the whole exome according to the same grouping rule, dividing the sample into a plurality of groups based on Panel according to the tumor mutation load value of Panel, and calculating the serious error grouping percentage according to the grade difference of the sample between the group based on the whole exome and the group based on Panel;

the sample grouping and deviation statistical test module is used for classifying the samples according to cancer species, cancer species and gender or cancer species and age, calling the Spearman correlation coefficient calculation module or the serious error grouping percentage calculation module, calculating the classified Spearman correlation coefficient or the serious error grouping percentage, and then carrying out statistical significance test on the influence of gender or age on consistency;

and the analysis module is used for analyzing the feasibility of estimating the tumor mutation load by the target sequencing Panel according to the Spearman correlation coefficient, the serious error grouping percentage and the statistical significance test result.

Preferably, the calculation formula of the tumor mutation load is as follows:

preferably, the grouping rule comprises evenly dividing the samples into no less than 3 levels according to the tumor mutation load value sorting.

Preferably, the calculation formula of the serious error grouping percentage Fs is as follows:

Fs＝S/(C+M+S)

wherein S is the number of samples grouped by gross errors, the gross errors being samples that differ by at least two degrees in the full exome-based group and in the Panel-based group;

c is the number of samples correctly grouped, with samples having the same rank in the full exome-based group and in the Panel-based group;

m is the number of samples grouped slightly in error, which is one level difference between samples in the full exome-based group and in the Panel-based group.

In a third aspect, the present invention provides a method of screening for a Panel gene to optimize identity, the method comprising:

calculating Spearman correlation coefficients for tumor mutation burden based on a single gene and tumor mutation burden based on a full exome;

screening genes from existing panels for sub-panels based on Spearman correlation coefficients and analyzing said sub-panels for the extent to which the feasibility of estimating tumor mutational burden is further improved;

genes were screened from the whole exome group based on Spearman correlation coefficient to form new panels, and the new panels were analyzed for the extent to which feasibility of estimating tumor mutational burden was further improved.

In a fourth aspect, the present invention provides an apparatus for screening for a Panel gene to optimize uniformity, the apparatus comprising:

a Spearman correlation coefficient calculation module for calculating a Spearman correlation coefficient for a tumor mutation burden based on a single gene and a tumor mutation burden based on a full exome;

a sub-Panel creation module for screening genes from existing Panel according to a Spearman correlation coefficient to form sub-Panel, and analyzing the sub-Panel to estimate the degree of improvement in the feasibility of tumor mutation burden;

a new Panel creation module for screening genes from the whole exome group based on the Spearman correlation coefficient to form a new Panel, and analyzing the new Panel to estimate the degree of increased feasibility of tumor mutational burden.

Compared with the prior art, the invention has the following beneficial effects:

(1) the consistency of TMB calculated by WES and TMB estimated by Panel is quantitatively evaluated by adopting error grouping percentage, and the result consistent with the result based on the correlation coefficient is obtained, so that the representativeness of Panel can be evaluated by taking the serious error grouping percentage as an index;

(2) the invention considers the poor representativeness of the commercial Panel in a specific cancer species, and the samples are further classified according to gender and/or age, which is beneficial to improving the representativeness of the Panel in a specific population of the specific cancer species;

(3) the method is favorable for obtaining the Panel with the best representativeness aiming at different purposes by calculating and comparing the Spearman correlation coefficient or the serious error grouping percentage of different Panels;

(4) according to the invention, through calculating the Spearman correlation coefficient or the serious error grouping percentage of a single gene, the first N genes (N is less than or equal to the total number of Panel genes) are selected to obtain sub-Panel with better representativeness, and the sub-Panel has an important reference value in the aspect of TMB evaluation; further, the present invention provides an implementation idea to select genes from all genes (not just from the existing Panel) to constitute a completely new Panel.

Drawings

FIG. 1 is a schematic flow chart of the feasibility of analyzing targeted sequencing Panel to estimate tumor mutation burden;

FIG. 2 shows the total number of samples and the distribution among 33 cancer species;

FIG. 3 is a graph of the correlation coefficient of Panel TMB with WES TMB when patients were grouped with only cancer species;

FIG. 4 illustrates a critical error packet decision rule;

FIG. 5 is the percentage of severe error groupings of Panel TMB and WES TMB when patients are grouped with only cancer species;

FIG. 6 is an assessment of the effect of gender on TMB compliance;

FIG. 7 is a graph of the percent severe error groupings of Panel TMB and WES TMB after classification of patients by cancer species and gender;

FIG. 8 is an evaluation of the effect of age on TMB consistency;

FIG. 9 is the percentage of severe error groupings of Panel TMB and WES TMB after classification of patients by cancer species and age;

fig. 10 is a comparison of correlation coefficients before and after Panel optimization.

Detailed Description

To further illustrate the technical means adopted by the present invention and the effects thereof, the present invention is further described below with reference to the embodiments and the accompanying drawings. It is to be understood that the specific embodiments described herein are merely illustrative of the invention and are not limiting of the invention.

The examples do not show the specific techniques or conditions, according to the technical or conditions described in the literature in the field, or according to the product specifications.

Example 1

The flow of analysis of feasibility of target sequencing Panel to estimate tumor mutation load is shown in FIG. 1:

calculating the tumor mutation load based on the full exome and the tumor mutation load based on the Panel according to the gene length and the mutation number of the full exome and the Panel respectively;

the calculation formula of the tumor mutation load is as follows:

calculating Spearman correlation coefficients for the full exome-based tumor mutation burden and the Panel-based tumor mutation burden;

uniformly dividing the sample into at least three grades based on the full exome according to the tumor mutation load value of the full exome according to the same grouping rule, similarly uniformly dividing the sample into a plurality of grades based on Panel and the same as the number of the grades of the full exome according to the tumor mutation load value of Panel, and calculating the serious error grouping percentage according to the grade difference of the sample in the group based on the full exome and the group based on Panel;

analyzing feasibility of estimating tumor mutation load by target sequencing Panel according to Spearman correlation coefficient and/or percentage of serious error grouping;

the calculation formula of the serious error grouping percentage is as follows:

Fs＝S/(C+M+S)

wherein S is the number of samples grouped by gross errors, the gross errors being samples that differ by at least two degrees in the full exome-based group and in the Panel-based group;

c is the number of samples correctly grouped, with samples having the same rank in the full exome-based group and in the Panel-based group;

m is the number of samples grouped slightly in error, which is one level difference between samples in the full exome-based group and in the Panel-based group.

Example 2

This example uses clinical data to analyze the feasibility of target sequencing Panel to estimate TMB, as follows:

(1) obtaining data

The data for all exome sequencing (WES) identified mutations were downloaded from the tcga (the Cancer Genome atlas) database and involved a total of 33 patients with 8706 Cancer species, the information being shown in figure 2;

downloading all gene structure information of the full exome based on RefSeq from a UCSC database for transcript merging and calculating the gene length;

downloading clinical information (including gender and age) of the sample from GDC (https:// GDC. cancer. gov/about-data/publications/pancreatalas);

a list of gene names for 11 commercial mainstream panels was obtained from published information. Panel has the code numbers FF, GP, MI, PSOC, TO, BROP, BOT, ITO, ITT, TT and G.

(2) Calculation of tumor mutational burden

Combining exons of genes with a plurality of transcripts in a whole exome according to exon-intron coordinates of the transcripts (combination principle: the site is an exon in any transcript, and the site is considered as an exon in the gene level), obtaining a union coordinate, summing all the lengths of the exons of the union of the genes to obtain the length of the genes, and summing all the lengths of the genes to obtain the length of the whole exome gene;

inquiring the length of each gene in Panel from the whole exome according to the gene name of Panel, and summing up to obtain the total length of all genes of Panel;

screening out mutations (including synonymous mutations and non-synonymous mutations) of the protein coding genes according to the mutation types and the gene name fields marked in the mutation data, and respectively calculating the number of the mutations of the protein coding genes in the whole exome and the Panel;

combining gene length and number of mutations, calculate the whole exome TMB (WES TMB) and Panel TMB as follows:

(3) calculating a correlation coefficient

Acquiring cancer species information of each patient from clinical information, and classifying the patients according to cancer types;

spearman correlation coefficients (Rs) were calculated for WES TMB and Panel TMB in 33 cancer species.

As shown in FIG. 3, it can be seen that when patients are grouped by cancer species alone, the correlation coefficient between Panel TMB and WES TMB is low in cancer species such as UVM, LGG, TGCT, PCPG, GBM, etc., indicating poor representativeness of Panel, which should not be used for immune medication decision.

(4) Calculating the percentage of severely erroneous packets

Patients were evenly divided into three groups, WES high (Top), medium (Middle) and low (Bottom), according to WES TMB size; similarly, patients were evenly divided into three groups, Panel high (Top), medium (Middle), and low (Bottom), according to Panel TMB size;

for a certain cancer patient, judging whether a serious error (serious False) group exists according to the group in the WES and the Panel, and if the group in the WES of the certain cancer patient is Top and the group in the Panel is Bottom, or if the group in the WES is Bottom and the group in the Panel is Top, judging that the patient is the serious error group, as shown in FIG. 4;

counting the number of severely error grouped patients (S) and the total number of patients (C + M + S) in 33 cancer species and 11 Panel, respectively, and calculating the percentage of severely error grouped patients (Fs) according to a formula;

Fs＝S/(C+M+S)

the results are shown in fig. 5, and it can be seen that when patients are grouped with cancer species alone, the percentage of severe error groupings of Panel TMB to WES TMB in cancer species such as UVM, LGG, TGCT, PCPG, GBM, etc. is high, indicating poor representation of Panel, consistent with the results of fig. 3 as a whole, indicating that the same assessment of Panel representation can be made using the percentage of severe error groupings as an index.

Example 3

This example evaluates the effect of gender on TMB consistency by the following steps:

the patients with 33 cancer species were further classified by sex, and of the 11 Panel of 33 cancer species, the male Fs and the female Fs were calculated, and the difference between the two was calculated, and the larger the difference was, indicating that the greater the difference in the percentage of serious errors grouping between men and women, the greater the influence of sex on TMB consistency. Fisher's Exact Test is adopted to Test the number of four types of patients with the right male grouping, the wrong male grouping, the right female grouping and the wrong female grouping, and the significant influence of the gender on the TMB consistency is analyzed. As shown in fig. 6, the results indicate that gender had a significant effect on TMB consistency of three cancer species, LGG, LIHC and LAML, but had substantially no effect on cancer species such as SKCM, lucc, LUAD.

The results of calculating Fs for gender-specific populations of cancer species are shown in fig. 7, and when patients are grouped by cancer species and gender, the percentage of serious errors in UVM women, PCPG men, GBM women, etc. is high, indicating that Panel is poor representativeness, and such populations should not use Panel for immune medication decision.

Example 4

This example evaluates the effect of age on TMB consistency by the following steps:

the age differences between the severe error group and the other groups of 33 cancer species and 11 Panel were calculated by classifying the patients of 33 cancer species into severe error (Seriously False) groups and other groups, respectively, and the larger the difference, the greater the age difference between the different groups, the greater the effect of age on TMB consistency. The age between groups was examined using T-Test to analyze the significant effect of age on TMB consistency. As shown in fig. 8, the results indicate that age has a significant effect on TMB consistency in most cancer species (e.g., THCA, LGG), while having substantially no effect on cancer species such as SKCM, UCEC, etc.

The Fs of a population of a particular age group (over 60 years defined as Old, and under 60 and 60 years defined as Young) in a particular cancer species were calculated and the results are shown in fig. 9, where patients were grouped with both cancer species and age, the percentage of serious errors in UVM age, PCPG age, TGCT youth, etc. was high, indicating poor representation of Panel, which should not be used for immune drug decision making.

Example 5

Calculating the TMB value of each gene aiming at twenty thousand protein coding genes in 33 cancer species, carrying out Spearman correlation coefficient calculation of the gene and WES TMB, and evaluating the TMB representing degree of each gene;

respectively sequencing 11 Panel genes according to the Spearman correlation coefficients of the genes, selecting the first N genes (N is an incremental value of 50, 100, 200 and … …, and N is not more than the total number of Panel genes) with the highest correlation coefficients to form sub-Panel, and calculating the Spearman correlation coefficients of sub-Panel TMB and WES TMB; the sub-Panel with the Spearman correlation coefficient being the maximum is taken as the optimized Panel, and the maximum is called the Optimal correlation coefficient (Optimal Rs).

As a result, as shown in fig. 10, it can be found that the optimized Panel correlation coefficients of 11 panels are all improved.

Similarly, for more than twenty thousand protein coding genes of the whole exome, the first N genes (N is less than or equal to the total number of the whole exome genes) with the highest correlation coefficient are selected to form a brand-new Panel. The larger N is, the higher the consistency is, but the more genes are required to be detected, so the cost is higher, and in practice, the value of N depends on the compromise balance between the pursuit of the correlation degree and the cost.

In conclusion, the consistency of TMB calculated by WES and TMB estimated by Panel is quantitatively evaluated by adopting the correlation coefficient and the percentage of serious error grouping, so that the effect of analyzing the feasibility of the TMB estimated by Panel is realized; classifying the samples according to cancer species and gender or age, improving the representativeness of Panel in a specific population of a specific cancer species; the method is beneficial to optimizing the existing Panel and constructing a brand-new Panel, and has important reference value in the aspect of TMB evaluation.

The applicant states that the present invention is illustrated in detail by the above examples, but the present invention is not limited to the above detailed methods, i.e. it is not meant that the present invention must rely on the above detailed methods for its implementation. It will be apparent to those skilled in the art that any modification, equivalent replacement of elements (mutation data, gene structure, cancer species, sex, age and Panel), selection of specific means, etc. of the product of the present invention falls within the scope and disclosure of the present invention.