Prediction method, system and storage medium for tumor purity and average ploidy information

Abstract

The present invention belongs to the field of information processing technology, and discloses a method, system, and storage medium for estimating tumor purity and average ploidy information. Read a standard sequence from a fasta file, read a sequencing read segment from a bam file; use BWA software to align the read segments, use SAMtools to extract the readcount value, define the length of Bin, remove useless positions and remove noise, and calculate the readdepth value of each bin that does not contain a missing position; use Free‑C software to detect the copy number variation CNV region; use a Gaussian mixture model to determine the type of loss region; determine the iteration range and step size; use a minimum deviation strategy to screen the optimal result, and determine that the tumor purity and average ploidy at this time are the final results. The present invention prevents the propagation of noise in the model, obtains a more reliable result, and shortens the amount of data that needs to be compared.

Description

Technical Field

[0001] The present invention belongs to the field of information processing technology, and in particular relates to a prediction method, system and storage medium for tumor purity and average ploidy information. Background Art

[0002] Gaussian mixture model is an algorithm for clustering data. Its basic assumption is that the overall distribution contains K sub-Gaussian distributions. The most important concepts of the entire algorithm are as follows: Gaussian distribution: also called normal distribution, its mathematical expectation μ determines the location of the distribution, and its standard deviation δ determines the amplitude of the distribution. Expected maximum algorithm: solves the likelihood equation in a model with latent variable probability parameters. This algorithm is used to iteratively solve the Gaussian mixture model. Maximum likelihood estimation: a common statistical method used to infer the parameters of the probability density function associated with a set of samples.

[0003] At present, there are mainly the following methods for detecting tumor purity and average ploidy based on next-generation sequencing technology: 1. Copy number-based method: estimate tumor purity by calculating the copy number of tumor and normal genomes; 2. SNV heterozygous site-based method: consider the mutation part of the alleles at the heterozygous sites with somatic mutations in the tumor sample to estimate the purity of the tumor, or estimate the ploidy of the tumor at the same time; 3. Method combining copy number changes and allele frequencies: combine copy number changes and allele frequencies at heterozygous sites in the normal genome to infer tumor purity and ploidy;

[0004] Among the above methods, the combination of copy number change methods can often achieve better accuracy in practice, and the method itself also shows stronger robustness. This is because the total number of reads in NGS data is very large, which gives this method a certain statistical basis. At present, the main ideas of using this method to detect tumor purity and ploidy can be divided into the following: (1) By determining the distribution of reads mapped to each part in the tumor region, a probability model is established to infer the purity and ploidy of the tumor. (2) Infer the purity and ploidy of the tumor by combining whole exome sequencing (WES) data. (3) Use read counts and combine certain prior knowledge to establish a model to infer tumor purity.

[0005] Through the above analysis, the problems and defects of the existing technology are as follows: the existing technology is sensitive to abnormal information in the detection of tumor purity and average ploidy, has excessive time complexity and long calculation time.

[0006] The difficulty of solving the above problems and defects is:

[0007] 1. It is difficult to ensure that the algorithm results can meet a certain accuracy while reducing the complexity of the algorithm. 2. It is a common phenomenon that data contains noise. When the noise is relatively large, it is difficult to ensure that the algorithm can still give an acceptable result.

[0008] The significance of solving the above problems and defects is that the algorithm has a faster execution speed and more reliable results, which can meet the needs of certain specific application scenarios. These application scenarios generally require the algorithm to give results as quickly as possible. Summary of the invention

[0009] In view of the problems existing in the prior art, the present invention provides a method, system and storage medium for predicting tumor purity and average ploidy information.

[0010] The present invention is implemented as follows: a method for predicting tumor purity and average ploidy information, the method for predicting tumor purity and average ploidy information comprising:

[0011] Read standard sequences from fasta files and read sequencing reads from bam files;

[0012] Use BWA software to align read segments, and use SAMtools to extract readcount values; define the length of Bin, remove useless positions and noise, and calculate the readdepth value of each bin that does not contain missing positions; use Free-C software to detect copy number variation CNV regions;

[0013] Use Gaussian mixture model to determine the type of loss area;

[0014] Determine the iteration range and step size;

[0015] The minimum deviation strategy is used to screen the optimal result. That is, when the deviation between the result calculated in a certain step and the actual observed value is less than the threshold, the tumor purity and average ploidy at this time are determined as the final result.

[0016] Further, the method of using the Gaussian mixture model to determine the type of the loss region includes: extracting a region of loss type [l1, l2, ..., l n ], and use the Gaussian mixture model to classify the loss region. Assume that there are two Gaussian distributions in the data, representing pure and heterozygous deletions, respectively, where the parameter of the jth Gaussian distribution is μ j , δ j ; For the i-th loss region, calculate its probability p under the j-th Gaussian distribution ij ; with p ijThe contribution of the i-th region to the j-th Gaussian distribution is used as the weight to calculate the weighted mean and variance. When the parameters of each Gaussian distribution converge, the final classification result is obtained, that is, label 0 represents pure and missing, and label 1 represents heterozygous missing.

[0017] Furthermore, determining the iteration range and step size includes: obtaining a set of classification labels label∈[0, 1], for pure and missing regions with label 0, using their labels to eliminate the average ploidy p in the target equation, and iterating the purity of the tumor; determining that the range of tumor purity is [0.05, 0.95], with a step size of 0.01; for heterozygous missing regions with label 1, substituting the calculated tumor purity, iterating the average ploidy using the target equation, and determining that the range of average ploidy is [1.7, 2.7], with a step size of 0.01.

[0018] Another object of the present invention is to provide a computer device, the computer device comprising a memory and a processor, the memory storing a computer program, and when the computer program is executed by the processor, the processor performs the following steps:

[0019] Read standard sequences from fasta files and read sequencing reads from bam files;

[0020] Use BWA software to align read segments, and use SAMtools to extract readcount values; define the length of Bin, remove useless positions and noise, and calculate the readdepth value of each bin that does not contain missing positions; use Free-C software to detect copy number variation CNV regions;

[0021] Use Gaussian mixture model to determine the type of loss area;

[0022] Determine the iteration range and step size;

[0023] The minimum deviation strategy is used to screen the optimal result. That is, when the deviation between the result calculated in a certain step and the actual observed value is less than the threshold, the tumor purity and average ploidy at this time are determined as the final result.

[0024] Another object of the present invention is to provide a computer-readable storage medium storing a computer program, wherein when the computer program is executed by a processor, the processor executes the following steps:

[0025] Read standard sequences from fasta files and read sequencing reads from bam files;

[0026] Use BWA software to align read segments, and use SAMtools to extract readcount values; define the length of Bin, remove useless positions and noise, and calculate the readdepth value of each bin that does not contain missing positions; use Free-C software to detect copy number variation CNV regions;

[0027] Use Gaussian mixture model to determine the type of loss area;

[0028] Determine the iteration range and step size;

[0029] The minimum deviation strategy is used to screen the optimal result. That is, when the deviation between the result calculated in a certain step and the actual observed value is less than the threshold, the tumor purity and average ploidy at this time are determined as the final result.

[0030] Another object of the present invention is to provide a system for predicting tumor purity and average ploidy information for implementing the method for predicting tumor purity and average ploidy information, wherein the system for predicting tumor purity and average ploidy information comprises:

[0031] Read data module, read standard sequence from fasta file, read sequencing read segment from bam file;

[0032] Determine CNV region module, used to detect copy number variation CNV region;

[0033] A module for determining the type of loss region, which is used to determine the type of loss region using a Gaussian mixture model;

[0034] Determine the iteration range and step size module, which is used to determine the search range of the algorithm;

[0035] The module for determining tumor purity and average ploidy is used to select the optimal result by adopting the minimum deviation strategy, that is, when the deviation between the result calculated in a certain step and the actual observed value is less than a threshold, the tumor purity and average ploidy at this time are determined as the final result.

[0036] Another object of the present invention is to provide a terminal equipped with the prediction system for tumor purity and average ploidy information.

[0037] Combining all the above technical solutions, the advantages and positive effects of the present invention are as follows: the present invention reduces the number of inferred variables, controls the mutual dependence between variables, and introduces correction values, which can prevent the propagation of noise in the model and obtain more reliable results; focuses on the loss part of the copy number variation region, greatly shortens the amount of data that needs to be compared, and can greatly improve the speed of the algorithm.

[0038] The present invention can control the mutual influence between variables and reduce the sensitivity of the model to abnormal values. At the same time, by reducing the area of ​​comparison, the model can obtain higher efficiency while ensuring the accuracy. BRIEF DESCRIPTION OF THE DRAWINGS

[0039] In order to more clearly illustrate the technical solutions of the embodiments of the present application, the following is a brief introduction to the drawings required for use in the embodiments of the present application. Obviously, the drawings described below are only some embodiments of the present application. For ordinary technicians in this field, other drawings can be obtained based on these drawings without creative work.

[0040] Figure 1 It is a flow chart of a method for predicting tumor purity and average ploidy information provided by an embodiment of the present invention.

[0041] Figure 2 is a schematic diagram of the structure of a prediction system for tumor purity and average ploidy information provided by an embodiment of the present invention;

[0042] Figure 2 In: 1. Data reading module; 2. CNV region determination module; 3. Loss region type determination module; 4. Iteration range and step size determination module; 5. Tumor purity and average ploidy determination module.

[0043] Figure 3 It is a flow chart of the implementation of the method for predicting tumor purity and average ploidy information provided by an embodiment of the present invention.

[0044] Figure 4 The embodiment of the present invention provides a schematic diagram of the results of comparing the present invention with the best known algorithm;

[0045] Figure 4 (a) is the tumor purity value output by each algorithm on the simulated sample; (b) represents the average ploidy value output by each algorithm on the simulated sample. DETAILED DESCRIPTION

[0046] In order to make the purpose, technical solution and advantages of the present invention more clearly understood, the present invention is further described in detail below in conjunction with the embodiments. It should be understood that the specific embodiments described herein are only used to explain the present invention and are not used to limit the present invention.

[0047] In view of the problems existing in the prior art, the present invention provides a method, system and storage medium for predicting tumor purity and average ploidy information. The present invention is described in detail below in conjunction with the accompanying drawings.

[0048] like Figure 1 As shown, the method for predicting tumor purity and average ploidy information provided by the present invention comprises the following steps:

[0049] S101: read standard sequences from fasta files and read sequencing reads from bam files;

[0050] S102: Use BWA software to align read segments, use SAMtools to extract readcount values; define the length of Bin, remove useless positions and noise, and calculate the readdepth value of each bin that does not contain a missing position; use Free-C software to detect copy number variation CNV regions;

[0051] S103: using a Gaussian mixture model to determine the type of the loss region;

[0052] S104: Determine the iteration range and step size;

[0053] S105: The minimum deviation strategy is used to select the optimal result, that is, when the deviation between the result calculated in a certain step and the actual observed value is less than a threshold, the tumor purity and average ploidy at this time are determined as the final result.

[0054] The method for predicting tumor purity and average ploidy information provided by the present invention can also be implemented by other steps by ordinary technicians in the industry. Figure 1 The method for predicting tumor purity and average ploidy information provided by the present invention is only a specific embodiment.

[0055] like Figure 2 As shown, the prediction system of tumor purity and average ploidy information provided by the present invention includes:

[0056] Read data module 1, read the standard sequence from the fasta file, and read the sequencing read segment from the bam file.

[0057] The CNV region determination module 2 is used to detect the copy number variation CNV region.

[0058] The module 3 for determining the type of the loss region is used to determine the type of the loss region using a Gaussian mixture model.

[0059] The module 4 for determining the iteration range and step size is used to determine the search range of the algorithm.

[0060] The module 5 for determining the tumor purity and average ploidy is used to select the optimal result by adopting the minimum deviation strategy, that is, when the deviation between the result calculated in a certain step and the actual observed value is less than a threshold, the tumor purity and average ploidy at this time are determined as the final result.

[0061] The technical solution of the present invention is further described below in conjunction with the accompanying drawings.

[0062] like Figure 3 As shown, the method for predicting tumor purity and average ploidy information provided by the present invention comprises the following steps:

[0063] (1) Data reading

[0064] Read standard sequences from fasta files and sequencing reads from bam files.

[0065] (2) Determine the CNV region

[0066] Use BWA software to align read segments, and use SAMtools to extract readcount values; define the length of the bin, remove useless positions and noise, and calculate the readdepth value of each bin that does not contain a missing position; use Free-C software to detect copy number variation CNV regions; (3)

[0067] (3) Using Gaussian mixture model to determine the type of loss area

[0068] From the CNV regions determined in (2), take out the region of loss type [l1, l2, ..., l n ], and use the Gaussian mixture model to classify the loss region. Assume that there are two Gaussian distributions in the data, which represent pure and heterozygous deletions, respectively, and the parameter of the jth Gaussian distribution is μ j , δ j For the i-th loss region, calculate its probability p under the j-th Gaussian distribution ij . ij The contribution of the i-th region to the j-th Gaussian distribution is used as the weight to calculate the weighted mean and variance. When the parameters of each Gaussian distribution converge, the final classification result can be obtained, that is, label 0 represents pure and missing, and label 1 represents heterozygous missing.

[0069] (4) Determine the iteration range and step size

[0070] Through the calculation of (3), the present invention obtains a set of classification labels label∈[0, 1]. For the pure and missing regions with label 0, the present invention uses its label to eliminate the average ploidy p in the target equation, thereby iterating the purity of the tumor. In the present invention, the present invention determines that the range of tumor purity is [0.05, 0.95], with a step size of 0.01. For the heterozygous missing region with label 1, the present invention substitutes the tumor purity calculated in the previous step, and at this time uses the target equation to iterate the average ploidy. The present invention determines that the range of average ploidy is [1.7, 2.7], with a step size of 0.01.

[0071] (5) Determine tumor purity and average ploidy

[0072] The minimum deviation strategy is used to screen the optimal result. That is, when the deviation between the result calculated in a certain step and the actual observed value is less than the threshold, the tumor purity and average ploidy at this time are determined as the final result.

[0073] The present invention studies the loss part of the copy number variation region, and reduces the calculation time of the model by reducing the area of ​​comparison; the present invention technically uses a Gaussian mixture model to classify the loss region, and by combining the characteristics of different missing regions, the purity and average ploidy of the tumor can be iteratively calculated;

[0074] The technical effects of the present invention are described in detail below in conjunction with simulation.

[0075] In order to verify the effectiveness of the present invention, the present invention is compared with the results of the best known algorithm on simulation data. Figure 4 As shown:

[0076] Figure 4 (a) is the value of tumor purity output by each algorithm on the simulated sample, where blue represents the value estimated from the groundtruth, which can be used as a reference standard. Brown represents the output value of the present invention, and dark green represents the output value of the existing algorithm ABSOLUTE. By comparing the present invention, it can be found that: 1. As the coverage increases, the output value of the present invention tends to be close to the estimated value; 2. The output result of the present invention is not much different from the output result of ABSOLUTE, which shows that the present invention can achieve an effect equivalent to that of the existing algorithm by reasonably reducing the comparison data; Figure 4 (b) represents the average ploidy value output by each algorithm on the simulated sample. The present invention can find that: 1. Except for the case where the coverage is 15x, the output value of the present invention has a small deviation from the estimated value. At other coverages, the method of the present invention approximates the average ploidy well; 2. In general, when ABSOLUTE is used to infer the average ploidy, a large deviation is produced.

[0077] It should be noted that the embodiments of the present invention can be implemented by hardware, software, or a combination of software and hardware. The hardware part can be implemented using dedicated logic; the software part can be stored in a memory and executed by an appropriate instruction execution system, such as a microprocessor or dedicated design hardware. It can be understood by a person of ordinary skill in the art that the above-mentioned devices and methods can be implemented using computer executable instructions and / or contained in a processor control code, such as a carrier medium such as a disk, CD or DVD-ROM, a programmable memory such as a read-only memory (firmware), or a data carrier such as an optical or electronic signal carrier. Such code is provided on the carrier medium. The device and its modules of the present invention can be implemented by hardware circuits such as very large-scale integrated circuits or gate arrays, semiconductors such as logic chips, transistors, etc., or programmable hardware devices such as field programmable gate arrays, programmable logic devices, etc., can also be implemented by software executed by various types of processors, and can also be implemented by a combination of the above-mentioned hardware circuits and software, such as firmware.

[0078] The above description is only a specific implementation mode of the present invention, but the protection scope of the present invention is not limited thereto. Any modifications, equivalent substitutions and improvements made by any technician familiar with the technical field within the technical scope disclosed by the present invention and within the spirit and principle of the present invention should be covered by the protection scope of the present invention.

Claims

1. A method for predicting tumor purity and average ploidy information, characterized in that: The method for predicting tumor purity and average ploidy information comprises: Read standard sequences from fasta files and read sequencing reads from bam files; Use BWA software to align read segments, use SAMtools to extract readcount values, define bin lengths, remove useless positions and noise, and calculate the readdepth value of each bin that does not contain a missing position; use Free-C software to detect copy number variation CNV regions; Use Gaussian mixture model to determine the type of loss area; Determine the iteration range and step size; The minimum deviation strategy is used to screen the optimal result. That is, when the deviation between the calculated result in a certain step and the actual observed value is less than the threshold, the tumor purity and average ploidy at this time are determined as the final result. The method of using the Gaussian mixture model to determine the type of the loss region includes: taking out a region of the loss type [l1, l2, ..., l n ], use the Gaussian mixture model to classify the loss area; assume that there are two Gaussian distributions in the data, representing pure and heterozygous deletions, respectively, where the parameter of the jth Gaussian distribution is μ j , δ j ; For the i-th loss region, calculate its probability p under the j-th Gaussian distribution ij ; with p ij As the contribution of the i-th region to the j-th Gaussian distribution, the weighted mean and variance are calculated. When the parameters of each Gaussian distribution converge, the final classification result is obtained, that is, label 0 represents pure and missing, and label 1 represents heterozygous missing; The determination of the iteration range and step size includes: obtaining a set of classification labels label∈[0, 1], for pure and missing regions with label 0, using their labels to eliminate the average ploidy p in the target equation, thereby iterating the purity of the tumor; determining that the range of tumor purity is [0.05, 0.95], with a step size of 0.01; for heterozygous missing regions with label 1, substituting the calculated tumor purity, iterating the average ploidy using the target equation, and determining that the range of average ploidy is [1.7, 2.7], with a step size of 0.

01.

2. A computer device, characterized in that: The computer device includes a memory and a processor, wherein the memory stores a computer program, and when the computer program is executed by the processor, the processor executes the steps of the method for predicting tumor purity and average ploidy information as claimed in claim 1.

3. A computer-readable storage medium storing a computer program, wherein when the computer program is executed by a processor, the processor executes the steps of the method for predicting tumor purity and average ploidy information as claimed in claim 1.

4. A system for predicting tumor purity and average ploidy information that implements the method for predicting tumor purity and average ploidy information according to claim 1, characterized in that: The prediction system of tumor purity and average ploidy information includes: Read data module, read standard sequence from fasta file, read sequencing read segment from bam file; Determine CNV region module, used to detect copy number variation CNV region; A module for determining the type of loss region is used to determine the type of loss region using a Gaussian mixture model, wherein determining the type of loss region using a Gaussian mixture model includes: extracting a region of loss type [l1, l2, ...l n ], use the Gaussian mixture model to classify the loss area; assume that there are two Gaussian distributions in the data, representing pure and heterozygous deletions, respectively, where the parameter of the jth Gaussian distribution is μ j , δ j ; For the i-th loss region, calculate its probability p under the j-th Gaussian distribution ij ; with p ij As the contribution of the i-th region to the j-th Gaussian distribution, the weighted mean and variance are calculated. When the parameters of each Gaussian distribution converge, the final classification result is obtained, that is, label 0 represents pure and missing, and label 1 represents heterozygous missing; The module for determining the iteration range and step size is used to determine the search range of the algorithm, wherein determining the iteration range and step size includes: obtaining a set of classification labels label∈[0, 1], for pure and missing regions with label 0, using their labels to eliminate the average ploidy p in the target equation, thereby iterating the purity of the tumor; determining that the range of tumor purity is [0.05, 0.95], with a step size of 0.01; for heterozygous missing regions with label 1, substituting the calculated tumor purity, using the target equation to iterate the average ploidy, determining that the range of average ploidy is [1.7, 2.7], with a step size of 0.01; The module for determining tumor purity and average ploidy is used to select the optimal result by adopting the minimum deviation strategy, that is, when the deviation between the result calculated in a certain step and the actual observed value is less than a threshold, the tumor purity and average ploidy at this time are determined as the final result.

5. A terminal, characterized in that: The terminal is equipped with the prediction system for tumor purity and average ploidy information as described in claim 4.

Images