A method and device for analyzing ctDNA low-frequency mutation sequencing data

Abstract

The invention discloses a low frequency mutation sequencing data analysis method and device. The method and the device are based on the iterative convergence of the frequency of a molecular tag to identify a clustering center, and accordingly the efficient and accurate clustering is realized. It is proved by analysis and test that the analysis result of a clustering strategy is more accurate thanthat of an ordinary clustering model, errors in PCR amplification and sequencing process can be eliminated, then the background noise problem in the sequencing analysis resulTS is solved, and the sensitivity and specificity of the detection are improved.

Description

technical field [0001] The invention belongs to the technical field of sequencing, and in particular relates to a method and device for analyzing ctDNA low-frequency mutation sequencing data. Background technique [0002] ctDNA (circulating tumor DNA), also known as circulating tumor DNA, is the DNA released by tumor cells into the blood circulation system, which can accurately reflect the molecular genetic information of the primary tumor tissue. By detecting ctDNA, point mutations, structural variations, and even chromosomes can be obtained. Genetic information such as copy number variation. Based on the heterogeneity of tumors, the content of ctDNA in the blood is very small, and the abundance of ctDNA in the early stage of cancer even reaches a level below 0.1%, which leads to a gene mutation frequency usually between 0.01% and 1% (low frequency mutation). [0003] Next-generation sequencing technology (NGS) is currently the most widely used sequencing technology, which...

AI Technical Summary

Problems solved by technology

[0003] Next-generation sequencing technology (NGS) is currently the most widely used sequencing technology, which has the advantages of high sequencing depth, high throughput, high accuracy, and good sensitivity. However, there are still technical difficulties in the application of NGS to the detection of low-frequency mutations in ctDNA

On the one hand, NGS inevitably has sequencing errors, and the single-base error rate is generally between 0.1% and 1%. -6 The replication error rate is about 10%, and it increases with the increase of the number of PCR cycles; these two factors lead to large background noise in the sequencing analysis of ctDNA low-frequency mutations, especially in the case of detection limits of 0.1% and below, the Difficulty distinguishing template DNA mutations from sequencing errors / replication errors, thus prone to false positive assay results

[0004] In order to solve the above problems, two methods are usually adopted: one is to increase the amount of sequencing data, and increase the sequencing depth by increasing the amount of data, so as to eliminate sequencing errors. However, the increase in data volume is not linearly related to the sequencing depth, and the error correction effect is very limited. ; The second is the molecular tag method, by connecting a molecular tag to at least one end of the original DNA molecule, the molecular tag can be a nucleotide sequence composed of random bases, the length is selected according to actual needs, based on the length of the molecular tag and the change of the base, In theory, molecular tags can have 4 n The original DNA molecules marked by molecular tags are unique; but in fact, there are still preference problems in the synthesis and labeling of DNA molecules with molecular tags, and there will inevitably be dominant molecular tags, and usually multiple DNA molecular tags In the case of the same molecular label, this increases the difficulty of identifying the original DNA molecule during the detection of ctDNA low-frequency mutations, and the detection accuracy is difficult to guarantee. Therefore, it is still necessary to improve the sequencing analysis method for ctDNA low-frequency mutations based on molecular labels. To improve the detection accuracy and avoid false positive results

Images