CN106676178A - System and method for tumor heterogeneity assessment - Google Patents

Abstract

The invention discloses a system and a method for tumor heterogeneity assessment and particularly provides a molecular clone analysis method. On the basis of various types of variation detection results of high-throughput sequencing in circulating tumor DNA, all variations are divided into different molecular clones to realize tumor heterogeneity assessment by molecular clone levels. By adoption of the method, tumor heterogeneity assessment based on ctDNA high-throughput variation detection is realized to effectively assist in making of tumor prognosis and treatment schemes.

Description

A kind of method and system of assessment Tumor Heterogeneity

Technical field

The invention belongs to biological technical field, particularly relates to assess the method for Tumor Heterogeneity and is System.

Background technology

Tumour is a kind of disease caused by gene alteration.Tumour often relates to various Genetic Variations types, including Single nucleotide variations (Single Nucleotide Variations, SNV), short insertion and deletion (small insertions And deletions, indel), copy number variation (Copy Number Variations, CNV), structure variation (Structure Variations, SV) etc..From the beginning of forming first variation from tumour cell, the mistake for accumulating that makes a variation just is started Journey.Favourable maintenance condition is built in propulsion and the evolution of tumour over time, the variation that the harmful variation for first occurring occurs after being, The ability for making tumour cell constantly obtain or strengthen the aspects such as suppression apoptosis, infinite copy, immunologic escape, therefore tumour The speed of the variation accumulation of cell is more faster than normal cell.The tumour for ultimately forming is in fact have different genetics characteristics The mixing of cell colony：Some cells only carry early stage variation, and some then also carry later stage variation simultaneously；In these tumour cells In, the involved cell proportion of variation also with its time of origin by morning to evening from large to small；Occur simultaneously in a cell Variation share a common destiny, life or death in tumour evolutionary process, the cell proportion being related to is identical.The cell proportion distribution of the variation in tumour Complexity can reflect Tumor Heterogeneity, and the latter is the most direct and most important embodiment of tumour complexity, and itself and tumour are suffered from Person's prognosis and life span are closely bound up.

The method for adopting the multidigit point sampling and high-flux sequence of same tumor patient assessment Tumor Heterogeneity at present more, i.e., Carry out after pathology sampling, by the method for high-flux sequence each being analyzed by the multiple positions of tissue to patient or multiple focuses The variation of sampling point, and be described and hierarchical statistics to having the corresponding cell proportion of variation and variation.The method has Following shortcoming：(1) there is bias in the clinical sampling in many sites, can only represent the molecular variant feature at taken position, it is impossible to generation The complexity of table entirety tumour；(2) with certain clinical risk；(3) MET of some types is difficult to obtain, such as pleuroperitoneum Metastatic lesion；(4) inaccuracy, by having the heterogeneous analysis method that variation is carried out, by total variation same layer is recognized as Level, is not specifically divided, so as to cause partial analysis result inaccurate (Gerlinger, M.et to total variation al.Intratumor heterogeneity and branched evolution revealed by multiregion sequencing.The New England journal of medicine 366,883-892,doi:10.1056/ NEJMoa1113205(2012)；Hao,J.J.et al.Spatial intratumoral heterogeneity and temporal clonal evolution in esophageal squamous cell carcinoma.Nature genetics,doi:10.1038/ng.3683(2016)).Additionally, also there is the copy number variation knot that method is only sampled by single-point Fruit is estimated (Oesper, L., Satas, G.＆Raphael, B.J.Quantifying tumor to Tumor Heterogeneity heterogeneity in whole-genome and whole-exome sequencing data.Bioinformatics 30,3532-3540,doi:10.1093/bioinformatics/btu651 (2014)), except there is sampling bias in the method Property shortcoming outside, also have the shortcomings that crowd's coverage is low, i.e., can only covering part exist a large amount of copy number variations cancerous swellings or Crowd.

Therefore, there is a need in the art for the heterogeneous analysis method of more accurate evaluation tumour, pre- with effectively adjuvant therapy Formulate with therapeutic scheme afterwards.

The content of the invention

For the heterogeneity of more accurate evaluation tumour, the present invention propose a kind of molecular cloning (Molecular Clone, MClone) analysis method, the method based on high-flux sequence Circulating tumor DNA (circulating tumor DNA, All variations are divided into different molecular clonings, using molecular cloning layer by the polytype variation testing result in ctDNA) The heterogeneity of level assessment tumour.The method of the present invention realizes that the Tumor Heterogeneity based on the variation detection of ctDNA high fluxs is commented Estimate, formulated with effective adjuvant therapy prognosis and therapeutic scheme.

Therefore, in a first aspect, the invention provides a kind of method of assessment Tumor Heterogeneity, methods described includes：

1) (preferred high-flux sequence) is sequenced to the dissociative DNA (cell-free DNA, cfDNA) of patient, is surveyed Sequence information；

2) determine that ctDNA makes a variation using the sequencing information, become according to the sequencing information and the ctDNA of the determination It is different, variation allelic frequency is calculated, actual total copy number of definitive variation region calculates the ratio that ctDNA accounts for cfDNA；

3) according to the step 2) in determine ratio and ctDNA variation sequencing information and copy number information to described CtDNA variations are clustered, and are clustered each cluster for obtaining and are defined as a molecular cloning, obtain the clone's level for clustering；

4) its Tumor Heterogeneity is estimated according to clone's level of the patient, clone's level of the patient is got over Many, its Tumor Heterogeneity is higher.

In second aspect, the invention provides a kind of heterogeneous method of the different patient tumors of comparison, methods described includes：

Using first aspect present invention method the step of 1) -3) calculate described each patient molecular cloning level, it is right In different patients, clone's level is more, and its Tumor Heterogeneity is higher.

In an embodiment of first or second aspect of the present invention, step 2) include：

2.1) variation V (the variation V is selected from SNV, indel and SV) (V is obtained using the sequencing information_i, i=1 ..., N) reference equipotential sequencing depth (R_i), variation equipotential sequencing depth (M_i), and calculate variation allelic frequency (Variant Allele Fraction, VAF_i),

Wherein, depth (R is sequenced with reference to equipotential_i) it is that the normal sequence of the variation does not occur in corresponding site in sequencing result Bar number；Variation equipotential sequencing depth (M_i) it is that the bar number of the series of variation of the variation occurs in corresponding site in sequencing result；

2.2) using variation V_iCNV (the CNV of region_i, i=1 ..., n), calculating variation V_iThe reference of region is copied Shellfish number (rCN_i) and actual total copy number (CN_i),

If in step 1) used in accurate CNV detection methods (such as using SNP chip detection), for not in male sex's property Variation on chromosome, can obtain the special copy number variation (CNV of equipotential on two chromosome_i,major, CNV_i,minor, CNV_i,major≥CNV_i,minor) information, the copy number (CN special so as to obtain actual equipotential_i,major, CN_i,minor),

2.3) ctDNA ratios assessment：With maximum variation allelic frequency to assess cfDNA in ctDNA proportions (CTF),

CTF=max (VAF_i), i=1 ..., n (formula 5)

In one embodiment, the step of method of the present invention 3) in, by the mutant ratio predicted, to variation Clustered, for example with PyClone (v0.13, such as current latest edition, non-specifically are indicated, and below refer both to the version) software.

In one embodiment, the step of method of the present invention 3) in, the reference and change of the V (SNV/indel/SV) that makes a variation Different equipotential depth data (R_i, M_i)：For assessing variation tumour cell ratio with mono- piece of CTF and CNV.In one embodiment, The step of method of the present invention 3) in, all tumour cells are accounted for variation place cell mass each described using PyClone softwares Ratio be predicted, software parameters can set as follows：Overall nodule cell proportion (CTF)=variation gene frequency Peak；Iterations=20000；Other specification is acquiescence.

In one embodiment, the step of method of the present invention 3) in, using PyClone to n the detected V that makes a variation (SNV/indel/SV) clustered, in addition to following parameter, using default parameters：

(a)--tumour_contentsCTF；

(b)--num_iters 20000；

(c) -- prior total_copy_number, when using the special CNV data of equipotential as input

When, the parameter is set to parental_copy_number；

(d) -- density pyclone_beta_binomial, when step 1) use the relatively low full base of sequencing depth During because of group sequencing technologies, the parameter is set to pyclone_binomial；

E () -- in_files patient.tsv, patient.tsv file is the file with tab as decollator；Remove Outside header line, often row includes the information of variation V (SNV/indel/SV)；Six row are included, is followed successively by：mutation_id、 Ref_counts, var_counts, normal_cn, minor_cn and major_cn.

In the third aspect, the invention provides a kind of system of assessment Tumor Heterogeneity, the system includes：

1) for the module of the cfDNA of (preferred high-flux sequence) patient to be sequenced；

2) for performing the module of following steps：

A) receive from module 1) sequencing information；

B) compared by the sequence information with normal gene sequence, obtain the ctDNA variations in cfDNA；

C) made a variation according to the sequencing information and the ctDNA, calculate variation allelic frequency, definitive variation region Actually the special copy number of the equipotential of total copy number or reality, calculates the ratio that ctDNA accounts for cfDNA；

D) sequencing information for being made a variation according to the ratio and ctDNA that determine in the step c) and copy number information are to described CtDNA variations are clustered, and determine molecular cloning, and calculate molecular cloning level；

3) result output module：

The result of Tumor Heterogeneity is exported according to the molecular cloning level of the patient, clone's level of the patient is got over Many, its Tumor Heterogeneity is higher.

In fourth aspect, the invention provides a kind of heterogeneous system of the different patient tumors of comparison, the system includes：

1) for the module of the cfDNA of (preferred high-flux sequence) patient to be sequenced；

2) for performing the module of following steps：

A) receive from module 1) sequencing information；

B) compared by the sequence information with normal gene sequence, obtain the ctDNA variations in cfDNA；

C) according to the sequencing information and the variation result, variation allelic frequency is calculated, definitive variation region Actually the special copy number of the equipotential of total copy number or reality, calculates the ratio that ctDNA accounts for cfDNA；

D) sequencing information for being made a variation according to the ratio and ctDNA that determine in the step c) and copy number information are to described CtDNA variations are clustered, and determine molecular cloning, and calculate molecular cloning level；

3) result output module：

The molecular cloning level of different patients is compared, the Tumor Heterogeneity result of the comparison of the different patients of output, Clone's level of patient is more, and its Tumor Heterogeneity is higher.

The present invention the 3rd or fourth aspect an embodiment in, module 2) the step of c) include step：

C.1) variation V (the variation V is selected from SNV, indel and SV) (V is obtained using the sequencing information_i, i=1 ..., N) reference equipotential sequencing depth (R_i), variation equipotential sequencing depth (M_i), and calculate variation allelic frequency (Variant Allele Fraction, VAF_i),

Wherein, depth (R is sequenced with reference to equipotential_i) it is that the normal sequence of the variation does not occur in corresponding site in sequencing result Bar number；Variation equipotential sequencing depth (M_i) it is that the bar number of the series of variation of the variation occurs in corresponding site in sequencing result；

C.2) using variation V_iCNV (the CNV of region_i, i=1 ..., n), calculating variation V_iThe reference of region is copied Shellfish number (rCN_i) and actual total copy number (CN_i),

If in step 1) used in accurate CNV detection methods (such as using SNP chip detection), for not in male sex's property Variation on chromosome, can obtain the special copy number variation (CNV of equipotential on two chromosome_i,major, CNV_i,minor, CNV_i,major≥CNV_i,minor) information, the copy number (CN special so as to obtain actual equipotential_i,major, CN_i,minor),

C.3) ctDNA ratios assessment：With maximum variation allelic frequency to assess cfDNA in ctDNA proportions (CTF),

CTF=max (VAF_i), i=1 ..., n (formula 5)

In an embodiment of the present invention the 3rd or fourth aspect, module 2) it is a plurality of instruction for performing the step Computer-readable medium.Module 3) be the instruction for performing the step computer-readable medium.

In one embodiment, the module 2 of system of the invention) the step of d) in, make a variation V's (SNV/indel/SV) With reference to and variation equipotential depth data (R_i, M_i)：For assessing variation tumour cell ratio with mono- piece of CTF and CNV.In a reality In applying scheme, the present invention system module 2) the step of d) in, using PyClone softwares to variation place cell each described Group accounts for the ratio of all tumour cells and is predicted, and software parameters can set as follows：Overall nodule cell proportion (CTF)=change The peak of heteroallele frequency；Iterations=20000；Other specification is acquiescence.

In one embodiment, the module 2 of system of the invention) the step of d) in, by the mutant ratio predicted Example, clusters, for example with PyClone softwares to variation.

In one embodiment, the module 2 of system of the invention) the step of d) in, using PyClone to the n that detects Individual variation V (SNV/indel/SV) is clustered, in addition to following parameter, using default parameters：

(a)--tumour_contentsCTF；

(b)--num_iters 20000；

(c) -- prior total_copy_number, when using the special CNV data of equipotential as input

When, the parameter is set to parental_copy_number；

(d) -- density pyclone_beta_binomial, when module 1) in use sequencing relatively low complete of depth During genomic sequencing technique, the parameter is set to pyclone_binomial；

E () -- in_files patient.tsv, patient.tsv file is the file with tab as decollator；Remove Outside header line, often row includes the information of variation V (SNV/indel/SV)；Six row are included, is followed successively by：mutation_id、 Ref_counts, var_counts, normal_cn, minor_cn and major_cn.

Theory and the high flux variation detection technique of ctDNA that the present invention is evolved based on tumour, swell from clone's level analysis Knurl makes a variation, there is provided more meet the heterogeneous appraisal procedure of tumor development rule.

It is a discovery of the invention that higher Tumor Heterogeneity has higher tumour progression risk.

Relative to other analysis methods, the advantage of the present invention is as follows：

1) information is comprehensive：Relative to unit point or the sample of tissue bias in many sites, ctDNA can react more fully Tumor cells feature；

2) convenience is sampled：Sample of tissue is typically derived from operation or punctures, compared to sample of tissue, especially many sites Sample of tissue, the detection of ctDNA only needs noninvasive blood sampling, clinically easily feasible；

3) high degree of accuracy：Make full use of variation information, cover SNV, indel and SV, retain variation concrete frequency and The bi-values that non-used detect/is not detected, based on tumour Evolution Theory, from clone's aspect, unmanifest aspect is carried out to heterogeneity Assessment.

3 points by more than, the method for the present invention and system can the more accurate and rational heterogeneities for assessing tumour.

Description of the drawings

By the following drawings, the present invention will be described.

Fig. 1 mClone analysis process figures, band * is that each patient is carried out respectively the step of indicate.

Fig. 2 survival analysises, left-most curve is high heterogeneous, and the right curve is low heterogeneity.

Specific embodiment

In the present invention, Gene Name is adopted using official's name (Official Symbol) in NCBI-Gene Representation generally in the art represents gene mutation and protein mutant.For example, c.518T>C (p.V173A) represents missense mutation, The T sequence changes that presentation code area is the 518th are C bases, so as to the amino acid for causing 173 sports smart ammonia by histidine V Sour A；C.2235_2249delGGAATTAAGAGAAGC (p.E746_A750del) represents small fragment disappearance, presentation code area the The bases G GAATTAAGAGAAGC disappearance of 2235 to 2249, so as to 5 amino acid for causing the 746th to 750 lack Lose；c.2663+1G>A represents shearing mutation, and the 2663rd place in presentation code area exon 3 end is close to first of introne Base changes into A by G；c.7081C>T (p.Q2361*) represents nonsense mutation, and the C sequence changes that code area is the 7081st are T alkali Base, so as to the Q for causing the 2361st is changed into terminator codon.

In the present invention, mathematic sign ceil is referred to and rounded up.

In the present invention, cfDNA can also be blood (blood plasma), saliva, pleural effusions and ascites, urine, excrement equal samples DNA。

In the present invention, the tumour is selected from, but not limited to,：Lung cancer, colorectal cancer, cancer of the stomach, breast cancer, kidney, pancreas Cancer, oophoroma, carcinoma of endometrium, thyroid cancer, cervical carcinoma, the cancer of the esophagus and liver cancer.It is described in a specific embodiment Tumour is lung cancer, and the variation is the variation listed in table 1.

Method of the present invention flow chart to ctDNA using high-flux sequence as shown in figure 1, to every tested patients, become It is different detected after, according to the sequencing result of ctDNA variations, assess ctDNA and account for the ratio of cfDNA；Aforementioned proportion and detection Variation together, as input, is clustered to variation, is clustered each cluster for obtaining and is defined as a molecular cloning, Ran Houji Clone's level is calculated, the Tumor Heterogeneity of every patient is estimated finally according to clone's level of all patients.The present inventor It was found that, for lung cancer, it is heterogeneous high patient more than 3.5 to clone level, and clone's level is low for heterogeneity less than 3.5 Patient.

It is below the main technical flows and principles and methods of the method for the present invention：

1. high-flux sequence detection ctDNA variations

First, to selecting several same cancer kind patients as study subject, row variation detection and parameter meter are entered to each patient Calculate：

1) by high throughput sequencing technologies and corresponding informatics such as the capture sequencings of full-length genome, full extron group or probe Analysis method, is sequenced to experimenter cfDNA, the variation included in ctDNA is obtained, including SNV, indel, SV, CNV etc.；

2) according to step 1) in sequencing result, obtain variation V (variation V be selected from SNV, indel and SV) (V_i, i= 1 ..., reference equipotential n) is sequenced depth (R_i), variation equipotential sequencing depth (M_i), and calculate variation allelic frequency (Variant Allele Fraction, VAF_i),

Wherein, depth (R is sequenced with reference to equipotential_i) it is that the normal sequence of the variation does not occur in corresponding site in sequencing result Bar number；Variation equipotential sequencing depth (M_i) it is that the bar number of the series of variation of the variation occurs in corresponding site in sequencing result；

3) using variation V_iCNV (the CNV of region_i, i=1 ..., n), calculating variation V_iThe reference copies of region Number (rCN_i) and actual total copy number (CN_i),

If the accurate CNV detection methods (such as being detected using SNP chip) used in 1), for not in the dyeing of male sex's property Variation on body, can obtain the special copy number variation (CNV of equipotential on two chromosome_i,major, CNV_i,minor, CNV_i,major ≥CNV_i,minor) information, the copy number (CN special so as to obtain actual equipotential_i,major, CN_i,minor),

Accurate CNV detections refer to the special copy number variation of the equipotential for obtaining two chromosomes, such as using SNP chip Detection.

2. variation cluster and clone's level are calculated

Then, to every patient, the variation for detecting is carried out cluster analysis and clone's layer by the parameter obtained in foundation 1 Level is calculated：

1) ctDNA ratios assessment：With maximum variation allelic frequency to assess cfDNA in ctDNA proportions (CTF),

CTF=max (VAF_i), i=1 ..., n (formula 5)

2) variation cluster：

For arbitrary variation (SNV/indel/SV), the derived cell of cfDNA is divided into three classes：Normal cell (N), The tumour cell (C0) for not carrying the variation is, the tumour cell (C1) for carrying the variation, carries the tumour cell (C1) of the variation The ratio referred to as variation tumour cell ratio of all tumour cells (C1+C0) is accounted for, if the variation tumour of two or more variations Cell proportion is suitable, then the time that they occur is approximate, can be endowed identical cluster label, is clustered into cluster, i.e., one point Son clone.

Therefore, the cluster that makes a variation is needed to use data below：

A) reference of variation V (SNV/indel/SV) and variation equipotential depth data (R_i, M_i)：For with CTF and CNV mono- Block assessment variation tumour cell ratio；

B) step 1.3) in reference copies number (rCN_i) and actual total copy number (CN_i) or the equipotential of reality special copy Shellfish number (CN_i,major, CN_i,minor)：For a certain variation, the copy number amplification of the variation equipotential or disappearance can cause the tumour that makes a variation The false rising or false reduction of cell proportion estimate, therefore add copy number delta data more accurately to judge C1 cells Genotype, correct variation frequency, correct assessment variation tumour cell ratio；

c)CTF：To the composition for estimating cfDNA derived cells, i.e., tumour cell (C0+ in all cells (N+C0+C1) C1 the ratio shared by), the accurate setting of the parameter contributes to being computed correctly from the reference equipotential of normal cell and from tumour The quantitative proportion of the reference equipotential of cell.

For example, n variation V (SNV/indel/SV) for detecting is entered using PyClone v0.13 (current latest edition) Row cluster, in addition to following parameter, using default parameters：

(a)--tumour_contentsCTF；

(b)--num_iters 20000；

(c) -- prior total_copy_number, when using the special CNV data of equipotential as input

When, the parameter is set to parental_copy_number；

D () -- density pyclone_beta_binomial, when 1.1) using the relatively low full genome of depth is sequenced During group sequencing technologies, the parameter is set to pyclone_binomial；

E () -- in_files patient.tsv, patient.tsv file is the file with tab as decollator；Remove Outside header line, often row includes the information of variation V (SNV/indel/SV)；Six row are included, is followed successively by：mutation_id、 Ref_counts, var_counts, normal_cn, minor_cn and major_cn.

PyClone(Roth,A.et al.PyClone:statistical inference of clonal

population structure in cancer.Nature methods 11,396-398,

doi:10.1038/nmeth.2883 (2014) .) V is estimated according to variation V (SNV/indel/SV) and CNV information_i The cell at place accounts for the ratio of all tumour cells, and according to this to each variation one cluster label (C of imparting_i, i=1 ..., n, C_i ∈ 1 ..., and c }, c is the number of cluster).

Other versions or other variation clustering softwares of PyClone can also be adopted to variation cluster.

3) clone level to calculate：

Clone's level, that is, make a variation the molecular cloning number c being polymerized to.Tumour during constantly progress, evolve and send out by tumour The structure of raw tree also gradually becomes huge and complexity, and molecular cloning also can be more, and clone's level is constantly deepened, therefore clones level Size it is closely related with Tumor Heterogeneity.

3. Tumor Heterogeneity assessment

Take clone's level median of all tested patients threshold value high/low as every patient tumors heterogeneity is judged； , less than the patient of the threshold value, its Tumor Heterogeneity is relatively low, otherwise then Tumor Heterogeneity is higher for clone's level.

Because genome mutation situation has notable difference, therefore the method for the present invention it is not recommended that across cancer kind ratio between cancer kind It is more heterogeneous.

In the method for the invention, in addition to sequencing steps, other steps can be in the form of instruction in calculating In machine computer-readable recording medium, it is only necessary to by the sequencing result input calculating equipment, the computing device can just read the calculating Instruction in machine computer-readable recording medium, completes other steps of the inventive method.The computing device includes but is not limited to computer, just Take formula computer, PAD, smart mobile phone, intelligent wrist etc..

Embodiment

In the present embodiment, by taking 10 patients with lung cancer as an example, the present invention will be described.It should be noted that the enforcement Example is merely to illustrate that purpose, and can not by any way be construed to the restriction to the application.

The variation list that 1.ctDNA high-flux sequences are detected

1) make a variation V (SNV/indel/SV)

10 patients with lung cancer detect respectively 2-8 variation, and variation V (SNV/indel/SV) detection list is shown in Table 1.

Variation V of table 1 (SNV/indel/SV) detection list

2)CNV

In 10 patients with lung cancer, only S5 detection EGFR amplifications, amplification times are 1.73, are shown in Table 2.Therefore, detect in S5 The EGFR Deletion total copy numbers of corresponding reality that make a variation be estimated as 4.

Table 2CNV detects list

Sample number into spectrum	Gene	Copy number variation state	Copy number variation multiple
				S5	EGFR	gain	1.73

2.mClone analysis results are counted

Pyclone is clustered

The variation for detecting is clustered using PyClone v0.13, in addition to following parameter, using acquiescence ginseng Number：

a)--tumour_contents

b)--num_iters 20000

c)--prior total_copy_number

d)--density pyclone_beta_binomial

e)--in_files

Parameter a) and e) respectively specify that CTF and input file.The CTF of each patient and the content of input file are shown in Table 3：

Table 3Pyclone input datas

Wherein, mutation_id represents variation numbering, and ref_counts represents reference count, and var_counts represents change Different counting, normal_cn represents normal copy number, i.e. CN_i, minor_cn represents little copy number, i.e. CN_i,minor, major_cn tables Show big copy number, i.e. CN_i,major。

The result and follow-up follow up data of the mClone analyses carried out using the method for the present invention as shown in table 4, is taken all The median of clone's level, i.e. cut-off=3.5, it is heterogeneous high patient to clone level more than 3.5, and clone's level is less than 3.5 are heterogeneous low patient.

Table 4mClone analysis results and the clinical information table of comparisons

Sample number into spectrum	Clone's level	Tumor Heterogeneity	Progression free survival phase (week)
				S1	2	It is low	54
S2	1	It is low	49
				S3	4	It is high	11
S4	4	It is high	27
				S5	6	It is high	9
S6	6	It is high	17
				S7	3	It is low	17
S8	3	It is low	34
				S9	5	It is high	22
S10	2	It is low	36

Survival analysis (see Fig. 2) is carried out to this batch of sample, the Tumor Heterogeneity result using clone's level assessment is found, it is right Patient's prognosis (progression free survival phase) has significant prediction effect, and (p, 0.044), higher Tumor Heterogeneity enters with higher Exhibition risk (Hazard ratio is 9.386).The result verification using mClone analytical technologies assess Tumor Heterogeneity validity and Accuracy.

The molecular cloning level that the molecular cloning mClone analysis methods of the present invention are obtained can be used for assessing the different of tumour Matter, and the heterogeneity of tumour represents the developing stage of tumour, in the heterogeneous bigger more later stage for representing patient in tumour, suffers from It is bigger that the tumour of person continues in the recent period developing risk.Above experimental data confirms this point.

Claims (11)

1. a kind of method of assessment Tumor Heterogeneity, methods described includes：

1) cfDNA of patient is sequenced (preferred high-flux sequence), is obtained sequencing information；

2) determine that ctDNA makes a variation using the sequencing information, made a variation according to the sequencing information and the ctDNA of the determination, meter Variation allelic frequency is calculated, actual total copy number of definitive variation region calculates the ratio that ctDNA accounts for cfDNA；

3) according to the step 2) in determine ratio and ctDNA variation sequencing information and copy number information to described CtDNA variations are clustered, and are clustered each cluster for obtaining and are defined as a molecular cloning, obtain the clone's level for clustering；

4) its Tumor Heterogeneity is estimated according to clone's level of the patient, clone's level of the patient is more, its Tumor Heterogeneity is higher.

2. a kind of heterogeneous method of the multiple patient tumors of comparison, methods described includes：

1) cfDNA of the plurality of Patient Sample A is sequenced (preferred high-flux sequence), is obtained sequencing information；

2) for each Patient Sample A, determine that ctDNA makes a variation using the sequencing information, according to the sequencing information and it is described really Fixed ctDNA variations, calculate variation allelic frequency, and actual total copy number of definitive variation region calculates ctDNA and accounts for The ratio of cfDNA；

3) according to the step 2) in determine ratio and ctDNA variation sequencing information and copy number information to described CtDNA variations are clustered, and are clustered each cluster for obtaining and are defined as a molecular cloning, obtain the cluster of each Patient Sample A Clone's level；

4) its Tumor Heterogeneity is compared according to clone's level of the plurality of patient, clones patient its tumour more than level It is heterogeneous high.

3. method according to claim 1 and 2, the step 2) include：

2.1) variation V is obtained using the sequencing information_i,=1 ..., reference equipotential sequencing depth R_i, variation equipotential sequencing it is deep Degree M_i, and calculate variation allelic frequency VAF_i,

${\mathrm{VAF}}_i=\frac{M_i}{M_i+R_i}*100\%,i=1...n$

Wherein, depth R is sequenced with reference to equipotential_iIt is that the bar number of the normal sequence of the variation does not occur in corresponding site in sequencing result； Variation equipotential sequencing depth M_iIt is that the bar number of the series of variation of the variation occurs in corresponding site in sequencing result；

2.2) using variation V_iThe CNV of region_i,=1 ..., calculate variation V_iThe reference copies number rCN of region_iAnd reality The total copy number CN in border_i,

${\mathrm{CN}}_i=\left\{\begin{array}{cc}ceil({\mathrm{CNV}}_i\&times;{\mathrm{rCN}}_i),& {\mathrm{CNV}}_i\&GreaterEqual;1\\ 1,& {\mathrm{CNV}}_i<1\end{array}\right.,i=1,...,n$

If in step 1) used in accurate CNV detection methods (such as being detected using SNP chip), for not in the dyeing of male sex's property Variation on body, can obtain the special copy number variation (CNV of equipotential on two chromosome_i,major, CNV_i,minor, CNV_i,major ≥CNV_i,minor) information, the copy number (CN special so as to obtain actual equipotential_i,major, CN_i,minor),

${\mathrm{CN}}_{i,\min or}=\left\{\begin{array}{cc}ceil\left({\mathrm{CNV}}_{i,\min or}\right),& {\mathrm{CNV}}_{i,\min or}\&GreaterEqual;1\\ 0,& {\mathrm{CNV}}_{i,\min or}<1\end{array}\right.,i=1,...,n$

${\mathrm{CN}}_{i,major}=\left\{\begin{array}{cc}ceil\left({\mathrm{CNV}}_{i,major}\right),& {\mathrm{CNV}}_{i,major}\&GreaterEqual;1\\ 1,& {\mathrm{CNV}}_{i,major}<1\end{array}\right.,i=1,...,n$

2.3) ctDNA ratios assessment：With the V that makes a variation_i,=1 ..., maximum variation allelic frequency to assess cfDNA in shared by ctDNA Ratio CTF,

CTF=max (VAF_i) ,=1 ...,.

4. method according to claim 1 and 2, the step 3) in by the mutant ratio of prediction, to make a variation into Row cluster, for example with PyClone softwares.

5. method according to claim 4, the variation V_i,=1 ..., reference and variation equipotential depth data R_iAnd M_i With mono- piece of assessment variation tumour cell ratio of CTF and CNV.

6. a kind of system of assessment Tumor Heterogeneity, the system includes：

1) for the module of the cfDNA of (preferred high-flux sequence) patient to be sequenced；

2) for performing the module of following steps：

A) receive from module 1) sequencing information；

B) compared by the sequence information with normal gene sequence, obtain the ctDNA variations in cfDNA；

C) made a variation according to the sequencing information and the ctDNA, calculate variation allelic frequency, the reality of definitive variation region Always the special copy number of the equipotential of copy number or reality, calculates the ratio that ctDNA accounts for cfDNA；

D) sequencing information for being made a variation according to the ratio and ctDNA that determine in the step c) and copy number information are to described CtDNA variations are clustered, and determine molecular cloning, and calculate molecular cloning level；

3) result output module：

The result of Tumor Heterogeneity is exported according to the molecular cloning level of the patient, clone's level of the patient is more, its Tumor Heterogeneity is higher.

7. a kind of heterogeneous system of the different patient tumors of comparison, the system includes：

1) for the module of the cfDNA of (preferred high-flux sequence) patient to be sequenced；

2) for performing the module of following steps：

A) receive from step 1) module sequencing information；

B) compared by the sequence information with normal gene sequence, obtain the ctDNA variations in cfDNA；

C) according to the sequencing information and the variation result, variation allelic frequency, the reality of definitive variation region are calculated Always the special copy number of the equipotential of copy number or reality, calculates the ratio that ctDNA accounts for cfDNA；

D) sequencing information for being made a variation according to the ratio and ctDNA that determine in the step c) and copy number information are to described CtDNA variations are clustered, and determine molecular cloning, and calculate molecular cloning level；

3) result output module：

The molecular cloning level of different patients is compared, the Tumor Heterogeneity result of the comparison of the different patients of output, patient Clone's level it is more, its Tumor Heterogeneity is higher.

8. the system according to claim 6 or 7, the module 2) the step of c) include step：

C.1) variation V is obtained using the sequencing information_i,=1 ..., reference equipotential sequencing depth R_i, variation equipotential sequencing it is deep Degree M_i, and calculate variation allelic frequency VAF_i,

${\mathrm{VAF}}_i=\frac{M_i}{M_i+R_i}*100\%,i=1...n$

Wherein, depth R is sequenced with reference to equipotential_iIt is that the bar number of the normal sequence of the variation does not occur in corresponding site in sequencing result； Variation equipotential sequencing depth M_iIt is that the bar number of the series of variation of the variation occurs in corresponding site in sequencing result；

C.2) using variation V_iThe CNV of region_i,=1 ..., calculate variation V_iThe reference copies number rCN of region_iAnd reality The total copy number CN in border_i,

${\mathrm{CN}}_i=\left\{\begin{array}{cc}ceil({\mathrm{CNV}}_i\&times;{\mathrm{rCN}}_i),& {\mathrm{CNV}}_i\&GreaterEqual;1\\ 1,& {\mathrm{CNV}}_i<1\end{array}\right.,i=1,...,n$

If in step 1) used in accurate CNV detection methods (such as being detected using SNP chip), for not in the dyeing of male sex's property Variation on body, can obtain the special copy number variation CNV of equipotential on two chromosome_i,majorAnd CNV_i,minorInformation, wherein CNV_i,major≥CNV_i,minor, the copy number CN special so as to obtain actual equipotential_i,majorAnd CN_i,minor,

${\mathrm{CN}}_{i,\min or}=\left\{\begin{array}{cc}ceil\left({\mathrm{CNV}}_{i,\min or}\right),& {\mathrm{CNV}}_{i,\min or}\&GreaterEqual;1\\ 0,& {\mathrm{CNV}}_{i,\min or}<1\end{array}\right.,i=1,...,n$

${\mathrm{CN}}_{i,major}=\left\{\begin{array}{cc}ceil\left({\mathrm{CNV}}_{i,major}\right),& {\mathrm{CNV}}_{i,major}\&GreaterEqual;1\\ 1,& {\mathrm{CNV}}_{i,major}<1\end{array}\right.,i=1,...,n$

C.3) ctDNA ratios assessment：With the V that makes a variation_i,=1 ..., maximum variation allelic frequency to assess cfDNA in ctDNA institutes Accounting example CTF,

CTF=max (VAF_i) ,=1 ...,.

9. the system according to claim 6 or 7, the module 2) and/or module 3) it is a plurality of finger for performing the step The computer-readable medium of order.

10. the system according to claim 6 or 7, the module 2) the step of d) in, make a variation V_i,=1 ..., reference and Variation equipotential depth data R_iAnd M_iWith mono- piece of assessment variation tumour cell ratio of CTF and CNV.

11. systems according to claim 6 or 7, the module 2) the step of d) in, by predict mutant ratio Example, clusters, for example with PyClone softwares to variation.