WO2023171859A1 - Method for distinguishing between somatic mutations and germline mutations - Google Patents

Abstract

The present invention relates to a method for distinguishing between somatic mutations and germline mutations. By the method, somatic mutations and germline mutations can be distinguished from each other using cell-free DNA that is relatively easy to obtain.

Description

How to Distinguish between Somatic and Germline Mutations

The present invention relates to a method for distinguishing between somatic mutations and germline mutations.

DNA mutations are the cause of cancer and are an important part of cancer research and treatment. Next-generation sequencing (NGS) is a promising technology for de novo mutation detection due to the vast number of reads that modern sequencers can generate. In theory, given sufficient read depth, all mutations or variations in a genomic sample can be observed, regardless of variant allele frequency (VAF) or genomic region.

However, it is not clear to reliably identify mutations due to noise within the reads. Several bioinformatic tools are being developed to identify mutations from sequencing reads.

Nevertheless, since germline mutations and somatic mutations cannot be distinguished through liquid biopsy in many cases, methods for detecting somatic mutations are continuously required.

One aspect of the present invention includes the steps of a) extracting cell-free DNA from a target sample and obtaining paired-end reads targeting cancer-related genes; b) deriving the size of the obtained pair-end read; and c) classifying somatic mutations and germline mutations from the fraction value of the short fragment size quantified among the derived pair-end reads. Distinguishing somatic mutations and germline mutations from cell-free nucleic acids. The purpose is to provide a method of providing information for.

One aspect of the present invention includes the steps of a) extracting cell-free DNA from a target sample and obtaining paired-end reads targeting cancer-related genes; b) deriving the size of the obtained pair-end read; and c) classifying somatic mutations and germline mutations from the fraction value of the quantified short fragment size among the derived pair-end reads; Provided is a method of providing information for distinguishing somatic mutations and germline mutations from cell-free nucleic acids containing a.

In one embodiment of the present invention, the target sample in step a) may be a sample isolated from a cancer patient.

In one embodiment of the present invention, the method of deriving the size of the obtained pair-end read in step b) is b-1) correcting the orientation of the mapped reads and calculating the difference in base sequence length of the plurality of reads. Calculating and obtaining a density plot according to fragment size; b-2) deriving a fragment value with a fragment size of 100 to 155 bp from the density plot; It may include.

In one embodiment of the present invention, the method for classifying somatic mutations and germline mutations in step c) is based on the fraction value of the quantified short fragment size among the derived pair-end reads and the overall average pair-end read. If the fraction value of the short fragment size quantified among pair-end reads is large by comparing the sizes, it may be classified as a somatic mutation.

The method of providing information for distinguishing between somatic mutations and germline mutations of the present invention can distinguish between somatic mutations and germline mutations using cell-free nucleic acids that are relatively easy to obtain.

Figure 1 is a graph showing the distribution of read fragment sizes from cfDNA samples of colon cancer patients according to allele frequency.

Figure 2 is a graph showing the distribution according to the fragment size of the read from the cfDNA sample of a colon cancer patient, divided into somatic mutations and germline mutations according to allele frequency.

Figure 3 shows an ROC curve created to distinguish germline mutations and somatic mutations according to various fragment sizes of reads derived from cfDNA samples of colon cancer patients.

Figure 4 is a box plotting graph that distinguishes germline mutations and somatic mutations using the average fragment size of reads derived from cfDNA samples of colon cancer patients.

Figure 5 is a box plotting graph distinguishing germline mutations and somatic mutations using AUC P1 of reads derived from cfDNA samples of colon cancer patients.

Figure 6 is a graph showing the distribution of read fragment sizes from cfDNA samples of lung cancer patients according to allele frequency.

Figure 7 is a graph showing the distribution according to the fragment size of the read from the cfDNA sample of a lung cancer patient, divided into somatic mutations and germline mutations according to allele frequency.

Figure 8 shows an ROC curve created to distinguish germline mutations and somatic mutations according to various fragment sizes of reads derived from cfDNA samples of lung cancer patients.

Figure 9 is a box plotting graph that distinguishes germline mutations and somatic mutations using the average fragment size of reads derived from cfDNA samples of lung cancer patients.

Figure 10 is a box plotting graph distinguishing germline mutations and somatic mutations using AUC P1 of reads derived from cfDNA samples of lung cancer patients.

One aspect of the present invention includes the steps of a) extracting cell-free DNA from a target sample and obtaining paired-end reads targeting cancer-related genes; b) deriving the size of the obtained pair-end read; and c) classifying somatic mutations and germline mutations from the fraction value of the short fragment size quantified among the derived pair-end reads. Distinguishing somatic mutations and germline mutations from cell-free nucleic acids. Provides a method of providing information for

The types of mutations that can be found in the human body largely include germline mutations and somatic mutations. Meanwhile, in the blood of cancer patients, circulating tumor DNA (ctDNA) and cell-free DNA (cfDNA) derived from the primary cancer are circulating together, and the present inventors used liquid biopsy to determine reproductive The present invention was completed by identifying a method for distinguishing between germline mutation and somatic mutation.

Step a) is a step of extracting cell-free DNA from the target sample and obtaining paired-end reads targeting cancer-related genes.

As used herein, the term 'sample' refers to a sample such as tissue, cells, whole blood, serum, plasma, saliva, sputum, cerebrospinal fluid, or urine from which a paired-end sequence can be obtained. It includes, but is not limited to, and may specifically include serum and plasma.

In one embodiment of the present invention, the target sample in step a) may be a sample isolated from a cancer patient.

The cancer may be any one of squamous cell carcinoma, adenocarcinoma, sarcoma, colon cancer, and lung cancer.

As used herein, the term 'sample' refers to a sample such as tissue, cells, whole blood, serum, plasma, saliva, sputum, cerebrospinal fluid, or urine from which a paired-end sequence can be obtained. It includes, but is not limited to, and may specifically include serum and plasma.

As used herein, 'cell-free DNA' or 'cfDNA' refers to a fragment of nucleic acid found outside of a cell (e.g., body fluid), and the body fluid is bloodstream, cerebrospinal fluid, saliva, or Including, but not limited to, urine. The cfDNA may be derived from the subject (e.g., from the subject's cells) or from a source other than the subject (e.g., from a viral infection).

In this specification, the 'paired-end sequence' refers to sequences cloned from both ends (pair-ends) of a gene of interest whose approximate distances from each other are known, and are sequenced in the forward and reverse directions. ) means one sequence. The method of obtaining the paired-end sequence may be performed by a method known in the art, and may preferably be obtained through next-generation sequencing (or 'NGS'). The specific method of next-generation sequencing technology is described in Metzker, M. (2010) Nature Biotechnology Reviews 11:31-46, which is incorporated herein by reference.

Step b) is a step of deriving the size of the obtained pair-end read. In one embodiment of the present invention, the method of deriving the size of the obtained pair-end read in step b) is b-1) Obtaining a density plot according to fragment size by correcting the orientation of the mapped reads and calculating the difference in base sequence length of the plurality of reads; b-2) It may include deriving a ratio value between the number of fragments with a fragment size of 100 to 155 bp and the number of fragments with a fragment size of 160 to 180 bp from the density plot.

To obtain a density plot according to fragment size by correcting the orientation of the mapped reads in step b-1) and calculating the difference in nucleotide sequence length of a plurality of reads, the method is as follows. Same as:

The pair-end sequence obtained through step a) is mapped to a reference read, and the size of the fragment is calculated using the coordinate value obtained after mapping. For example, if read1 is mapped to chr1:100-250 and read2 is mapped to chr1:100-250 in a specific read fragment, after correcting the orientation of the read fragment, read2 The difference between chr1:250, the end, and chr1:100, the end of read1, which is 150 bp, can be obtained. If 100 of the reads are mapped, the fragment size can be calculated for the 100 reads using the above calculation method. For example, 90 bp fragments can be classified into 10 fragments, 105 bp fragments are 20, 130 bp fragments are 30, and 160 bp fragments are 40, and a density plot can be created through this. there is.

According to one embodiment of the present invention, after step b-1), step b-2) is performed to derive a fragment value with a fragment size of 100 to 155 bp from the density plot.

For example, if the AUC (area under curve) value for 100 to 155 bp (AUC P1) and the AUC value (AUC P2) for 160 to 180 bp are calculated according to the number of fragment sizes for the 100 reads, AUC P1 can be 10+20+30/100 = 0.6, AUC P2 can be 40/100 = 0.4, and the overall average fragment size can be calculated using the genomic adjustment value of the read fragment.

Step c) is a step of classifying somatic mutations and germline mutations. In one embodiment of the present invention, the method of classifying somatic mutations and germline mutations of step c) includes quantification of the derived pair-end reads. By comparing the fraction value of the short fragment size with the size of the overall average pair-end read, if the fraction value of the quantified short fragment size among the pair-end reads is large, it may be classified as a somatic mutation. .

Hereinafter, one or more specific examples will be described in more detail through examples. However, these examples are intended to illustrate one or more embodiments and the scope of the present invention is not limited to these examples.

Example 1: Method for deriving fragment distribution of pair-end sequence in cfDNA

cfDNA was obtained from blood samples obtained from 322 colon cancer patients and 100 lung cancer patients who visited the Department of Hematology and Oncology at Seoul National University Hospital using Promega's Maxwell automated equipment (Maxwell® RSC ccfDNA Plasma Kit) according to the manufacturer's protocol. Afterwards, NGS (Next Generation Sequencing) was performed on the obtained cfDNA. When performing targeted panel sequencing, the NGS DNA library prep kit (IMBdx) was used, and the target region was amplified using AlphaLiquid® 100 target capture panel (IMBdx). 150bp pair-end sequencing was performed using Illumina's NextSeq 550 platform.

Afterwards, the obtained pair-end sequencing results were aligned using the Burrows-Wheeler Aligner (BWA, version 0.7.10) “mem” algorithm. In this example, hg38, a known human gemone, was used as a reference genome. For variant calling, IMBdx's UniqSeq protocol was used and the company's in-house filtering steps were used (by default, the variant caller used is vardict).

The size of the fragment was determined using the genomic coordinate of the read fragment generated after alignment. For example, if read1 is mapped to chr1:100-250 and read2 is mapped to chr1:100-250 in a specific read fragment, after correcting the orientation of the read fragment, read2 The difference between chr1:250, the end, and chr1:100, the end of read1, which is 150 bp, can be obtained. The fragments were classified into 10 90 bp fragments, 20 fragments of 105 bp, 30 fragments of 130 bp, and 40 fragments of 160 bp, and a density plot was created from the calculated DNA fragment sizes. The actual fragment size calculated to create the density plot was calculated as a continuous value of 80 to 1000 bp, and the number was counted accordingly.

From the density plot, AUC P1 (100-155 bp) and AUC P2 (160-180 bp) values were calculated. From the number of classified segments, AUC P1 can be derived as 10+20+30 / 100 = 0.6, and AUC P2 = 40 / 100 = 0.4. In addition, the overall average fragment size was calculated using the genomic adjustment value of the read fragment generated after the alignment, and the calculated AUC P1 value and the average fragment size were used to determine germline mutation and somatic mutation. It was used as a reference indicator to distinguish between mutations.

Example 2: Distinction between germline mutations and somatic mutations according to section size from colon cancer patient samples

The distribution of the fragment sizes of the leads derived according to the method of Example 1 for plasma cfDNA samples from 322 colorectal cancer patients was plotted according to the allele frequency (Variant Allele Frequency, VAF). Figure 1 is a graph showing the size of the actual reads according to the VAF value by dividing the alteration allele and reference allele according to each mutation type, and through this, somatic mutation. In the case of , it was confirmed that reads with variant alleles were distributed with shorter fragment sizes even at low VAF, and the diagram in Figure 1 was classified into somatic mutations and germline mutations and subdivided each into somatic mutations and germline mutations. As a result, as shown in FIG. 2, it was confirmed that the lead fragment that actually had the modified allele of the somatic mutation had a shorter fragment size than the lead fragment that was confirmed to be a germline mutation.

Figure 3 shows the results of confirming through the ROC curve created according to the method of Example 1 that germline mutations and somatic mutations that do not exist in the existing database can be distinguished using values obtained from various fragment sizes. AUC P1 is shown in dark blue, and the average intercept size is shown in light orange. From the above results, it was confirmed that when using the AUC P1 value and average fragment size, somatic mutations and germline mutations can be distinguished with high accuracy.

Figures 4 and 5 show box plotting results distinguishing germline mutations and somatic mutations using the average intercept and AUC P1 derived according to the method of Example 1. As shown in Figures 4 and 5, in the case of a germline variant that is not registered in the germline variant database (e.g., GNOMAD, etc.) most commonly used by existing researchers, the AUC P1 value and average fragment size are used. It was confirmed that somatic mutations and germline mutations were statistically significantly distinguished.

Example 3: Distinction between germline mutations and somatic mutations according to section size from lung cancer patient samples

The distribution of the fragment sizes of the leads derived according to the method of Example 1 for plasma cfDNA samples from 100 lung cancer patients was plotted according to the allele frequency (Variant Allele Frequency, VAF). Figure 6 is a graph showing the size of the actual reads according to the VAF value by dividing the alteration allele and reference allele according to each mutation type, and through this, somatic mutation. In the case of , it was confirmed that reads with variant alleles were distributed with shorter fragment sizes even at low VAF, and the diagram in Figure 6 was classified into somatic mutations and germline mutations and subdivided each into somatic mutations and germline mutations. As a result of confirmation, as shown in Figure 7. In fact, it was confirmed that the leads with more variant alleles of somatic mutations had shorter fragment sizes than the lead fragments confirmed to be germline mutations.

Figure 8 shows the results of confirming through the ROC curve created according to the method of Example 1 that germline mutations and somatic mutations that do not exist in the existing database can be distinguished using values obtained from various fragment sizes. AUC P1 is shown in dark blue, and the average intercept size is shown in light orange. From the above results, it was confirmed that when using the AUC P1 value and average fragment size, somatic mutations and germline mutations can be distinguished with high accuracy.

Figures 9 and 10 show box plotting results for distinguishing germline mutations and somatic mutations using the average intercept and AUC P1 derived according to the method of Example 1. As shown in Figures 9 and 10, in the case of a germline variant that is not registered in the germline variant database (e.g., GNOMAD, etc.) most commonly used by existing researchers, the AUC P1 value and average fragment size are used. It was confirmed that somatic mutations and germline mutations were statistically significantly distinguished.

So far, the present invention has been examined focusing on its preferred embodiments. A person skilled in the art to which the present invention pertains will understand that the present invention may be implemented in a modified form without departing from the essential characteristics of the present invention. Therefore, the disclosed embodiments should be considered from an illustrative rather than a restrictive perspective. The scope of the present invention is indicated in the claims rather than the foregoing description, and all differences within the equivalent scope should be construed as being included in the present invention.

Claims (4)

a) extracting cell-free DNA from a target sample to obtain paired-end reads targeting cancer-related genes; b) deriving the size of the obtained pair-end read; and c) classifying somatic mutations and germline mutations from the fraction value of the quantified short fragment size among the derived pair-end reads; A method of providing information for distinguishing between somatic and germline mutations from cell-free nucleic acids, comprising: According to paragraph 1, A method in which the target sample in step a) is a sample isolated from a cancer patient. According to paragraph 1, The method for deriving the size of the obtained pair-end lead in step b) is b-1) correcting the orientation of the mapped reads and calculating the difference in base sequence length of a plurality of reads to obtain a density plot according to fragment size; b-2) deriving a fragment value with a fragment size of 100 to 155 bp from the density plot; A method comprising: According to paragraph 1, The method for classifying somatic mutations and germline mutations in step c) is By comparing the fraction value of the quantified short fragment size among the derived pair-end reads and the size of the overall average pair-end read, the fraction value of the quantified short fragment size among the pair-end reads was determined. In large cases, the method is to classify it as a somatic mutation.

Images