Disease Early Detection Device and Method Based on Deep Cell Free Nucleic Acid Sequencing

Abstract

The present invention relates to a disease early detection device based on deep cell-free nucleic acid sequencing, comprising: a library production unit that produces a library set including a full-length genome sequencing library, a full-length genome epigenetic sequencing library, a transcriptome sequencing library, and an epigenetic transcriptome sequencing library based on cell-free nucleic acids extracted from blood; a sequencing unit that sequences the library set to generate deep cell-free nucleic acid sequencing data; a signal calculation unit that maps the data to a genome reference to identify its position on the genome and calculates at least one disease-specific signal value based on the mapped data; a prediction signal value generation unit that processes at least one disease-specific signal value to generate a disease prediction signal value; and a prediction model generation unit that generates a disease prediction model that detects the presence or absence of disease and the type of disease in the input blood using labeling information related to blood and the disease prediction signal value.

Description

[Technical Field]

[0001] This invention relates to a disease detection technology based on deep nucleic acid sequencing, and more particularly to a technology that can detect diseases early based on deep sequencing of cell-free nucleic acids. [Background technology]

[0002] Even in modern times, with its advanced medicine, the majority of human deaths are caused by disease. According to 2019 statistics, 88% of global human deaths were due to disease. More specifically, non-communicable diseases accounted for 74%, infectious diseases for 14%, and non-communicable diseases such as cancer and heart disease accounted for 51% of human deaths. https: / / ourworldindata.org ).

[0003] One of the main causes of death from disease is the inability to recognize the onset of the disease early. Taking cancer, which is one of the leading causes of death worldwide, as an example, cancer is often discovered at a fairly advanced stage, and the majority of patients are very difficult to treat, with many dying within two years. Such high cancer mortality rates are closely related not only to problems with treatment methods but also to the lack of methods for early detection of cancer.

[0004] Early detection of disease has been consistently reported to significantly increase patient survival rates. According to a 2022 report by Cancer Research UK, there is a very large difference in survival rates between early and late-stage cancer detection. Survival rates for lung, colorectal, and breast cancer were high—approximately 60%, 90%, and 100% respectively—when detected early, while they were extremely low—approximately 10%, 10%, and 30% respectively—when detected late. This is especially true for pancreatic cancer, where symptoms are often minimal in the early stages and effective early detection methods do not exist; in many cases, cancer is only detected in the late stages, often leading to death.

[0005] While it can vary depending on the type of disease, early detection is quite difficult. Currently, for a disease to be detected early, patients need to voluntarily undergo thorough health checkups for multiple diseases at an early stage (even if there is no pain or only mild symptoms), and even then, a disease is not always detected. Taking cancer as an example, current clinical practice begins with a medical history and physical examination, followed by radiography and endoscopy if suspicious findings are present, and finally, cancer is confirmed by tissue examination. However, for cancer to be detected by clinical testing methods, a considerable number of cancer cells must be present and the diameter of the tumor must be above a certain size. Cancers of this size are generally discovered in more than half of cases after metastasis. Furthermore, current clinical cancer detection technology is limited to detecting specific types of cancer, and the need to perform multiple independent cancer-specific tests makes cancer detection difficult and cumbersome. Very few early-stage cancer patients with no or very few symptoms undergo these numerous independent cancer detection tests preventively and voluntarily.

[0006] Developing easily accessible and convenient disease detection tests, such as blood tests, could increase the frequency of voluntary disease screenings and improve the sensitivity of early disease detection. Numerous previous studies suggest that certain diseases can be detected by analyzing sequencing data of cell-free nucleic acids in the blood. Such attempts have been made for a variety of diseases, most notably cancers such as colorectal cancer, lung cancer, and breast cancer (Klein et al., 2022, Ann. Oncol.; Fairchild et al., 2023, Sci. Transl. Med.), as well as rare diseases including Erdheim-Chester disease and Klippel-Trenaunay syndrome (Janku et al, 2014, Oncotarget; Palmieri et al, 2020, Vascular.), and immune diseases including systemic lupus erythematosus and rheumatoid arthritis (Truszewska et al., 2020, Lupus; Ding et al., 2022, Clin. Chem.; Zhou et al., 2023, PNAS; Plant et al., 2016, These include chronic diseases such as arthritis (Rheumatol.), hypertension (Brusca et al., 2022, Circulation), cardiovascular diseases such as chronic heart disease (Zemmour et al., 2018, Nat. Commun.), metabolic diseases such as fatty liver disease and diabetes (Karlas et al., 2017, J. Transl. Med.; Lehmann-Werman et al., 2016, PNAS), and infectious diseases such as infectious pneumonia and invasive fungal infections (Langelier et al., 2020, Am. J. Respir. Crit. Car. Med.; Hong et al., 2018, Diagn. Microbiol. Infect. Dis.).

[0007] Recently, attempts have been reported to extract disease detection markers from sequencing data of cell-free nucleic acids in patients' blood to detect diseases early. In a recent cancer study (Klein et al., 2022, Ann. Oncol.), panel sequencing of an epigenomic library of cell-free nucleic acids was reported to be able to detect stage 1 cancer with a sensitivity of approximately 17% and a specificity of 99.5%. However, this is considerably lower than the detection sensitivity for stage 3 and 4 cancer in the same study (approximately 70% and a specificity of 99.5%). Another study (Nicholson et al., 2023, Lancet) also reported a low detection rate of approximately 24% and a specificity of 98% for stage 1 cancer. While these results from previous studies suggest the possibility of early disease detection, they also strongly indicate the need for the development of highly accurate early disease detection tests.

[0008] On the other hand, recent studies have reported the early detection of diseases using full-length genome sequencing rather than panel sequencing, sequencing free nucleic acids from cells in the blood at low sequencing depths of 1 to 5 (Cristiano et al., 2019, Nature; Mouliere et al., 2019, Sci. Transl. Med.; Wan et al., 2022, Nat. Commun.; An et al., 2023, Nat. Commun.; Bruhm et al., 2023, Nat. Genet.; Budhraja et al., 2023, Sci. Transl. Med.). Depth refers to the average number of sequencing reads that have specific location information on the genome. Higher depths increase the number of sequencing reads corresponding to specific locations on the genome, potentially leading to a greater quantity and accuracy of genomic information obtained. This suggests that detecting diseases using deep cell-free nucleic acid sequencing may offer the potential for more accurate early detection compared to conventional cell-free nucleic acid-based cancer detection methods. Therefore, there is a need for methods to detect diseases early based on such deep cell-free nucleic acid sequencing. [Overview of the Initiative] [Problems that the invention aims to solve]

[0009] One embodiment of the present invention relates to a disease early detection device and method based on deep cell-free nucleic acid sequencing, which can obtain deep cell-free nucleic acid sequencing data of cell-free nucleic acids extracted from blood.

[0010] One embodiment of the present invention relates to a disease early detection device and method based on deep cell-free nucleic acid sequencing, which can detect diseases by calculating disease-specific signal values ​​based on deep cell-free nucleic acid sequencing data mapped to a genome reference.

[0011] One embodiment of the present invention relates to a disease early detection device and method based on deep cell free nucleic acid sequencing, which can detect the presence or absence and type of disease in a person examined solely from cell free nucleic acids extracted from the blood of the person examined, using disease prediction signal values ​​obtained by deep sequencing of free nucleic acids extracted from blood.

[0012] One embodiment of the present invention relates to a method for early detection of disease using only free nucleic acids extracted from a subject's blood, by using disease prediction signal values ​​obtained by high-depth sequencing of free nucleic acids extracted from blood. [Means for solving the problem]

[0013] The present invention relates to a disease detection device based on deep cell-free nucleic acid sequencing according to one embodiment of the present invention, comprising: a library production unit that produces a library set including a full-length genome sequencing library, a full-length genome epigenetic sequencing library, a transcriptome sequencing library and a transcriptome epigenetic sequencing library based on cell-free nucleic acids extracted from blood; a sequencing unit that generates deep cell-free nucleic acid sequencing data by sequencing the library set at a high depth of sequencing depth of 10 or more; a signal calculation unit that maps the deep cell-free nucleic acid sequencing data to a genome reference to identify the position of the data on the genome and calculates at least one disease-specific signal value based on the mapped data; a prediction signal value generation unit that processes the at least one disease-specific signal value to generate a disease prediction signal value; and a prediction model generation unit that generates a disease prediction model that detects the presence or absence and type of disease in the input blood using labeling information related to blood and the disease prediction signal value.

[0014] The blood samples can be collected individually from individuals without disease, individuals with a history of a specific disease but who are currently disease-free, patients in the early, middle, and late stages of a specific disease, and patients whose specific disease has relapsed.

[0015] The signal calculation unit can calculate the disease-specific signal value based on one or more of the DNA signal value, DNA deformation signal value, RNA signal value, and RNA deformation signal value.

[0016] The predictive signal value generation unit can process multiple disease-specific signal values ​​as vector data and generate the disease predictive signal value.

[0017] The aforementioned prediction signal value generation unit can generate disease prediction signal values ​​according to the type of target disease.

[0018] The aforementioned high-depth cell-free nucleic acid sequencing-based disease early detection device may further include a disease prediction unit that receives blood data extracted from a person being tested as input and detects whether or not the person being tested has developed a disease and the type of disease through the disease prediction model.

[0019] The sequencing unit can use multiple sequencers to sequence the library set.

[0020] A disease detection method based on high-depth cell-free nucleic acid sequencing according to an embodiment of the present invention, comprising the steps of: preparing a library set including a full-length genome sequencing library, a full-length epigenetic genome sequencing library, and a transcriptome sequencing library based on cell-free nucleic acids extracted from blood; sequencing the library set to generate high-depth cell-free nucleic acid sequencing data; mapping the high-depth cell-free nucleic acid sequencing data to a genome reference to identify the genomic position of the high-depth cell-free nucleic acid sequencing data, and calculating at least one disease-specific signal value based on the mapped high-depth cell-free nucleic acid sequencing data; processing the at least one disease-specific signal value to generate a disease prediction signal value; generating a disease prediction model using the labeling information related to the blood and the disease prediction signal value; and detecting the presence or absence of disease onset and the type of disease in a test subject using the disease prediction model and blood data extracted from the test subject. [Advantageous Effects of the Invention]

[0021] The technology of the present disclosure can achieve the following advantageous effects. However, this does not mean that a specific embodiment has all of the effects shown below, or has only the following effects, so the scope of rights of the technology of the present disclosure should not be limited thereby.

[0022] The apparatus and method for detecting diseases based on high-depth cell-free nucleic acid sequencing according to an embodiment of the present invention can obtain high-depth cell-free nucleic acid sequencing data for cell-free nucleic acids extracted from blood.

[0023] The apparatus and method for detecting diseases based on high-depth cell-free nucleic acid sequencing according to an embodiment of the present invention is an apparatus and method capable of calculating disease-specific signal values based on high-depth cell-free nucleic acid sequencing data mapped to a genome reference, so as to detect various diseases at an early stage.

[0024] An apparatus and method for detecting a disease based on high-depth cell-free nucleic acid sequencing according to an embodiment of the present invention can early detect the presence or absence of the onset of a disease in a subject using only cell-free nucleic acids contained in the blood of the subject, by using a disease prediction signal value obtained by performing high-depth sequencing on the cell-free nucleic acids extracted from the blood.

[0025] An apparatus and method for detecting a disease based on high-depth cell-free nucleic acid sequencing according to an embodiment of the present invention can early identify the type of a disease using only cell-free nucleic acids contained in the blood of a subject, by using a disease prediction signal value obtained by high-depth sequencing of cell-free nucleic acids extracted from blood. [Brief Description of the Drawings]

[0026] [Figure 1] Figure 1 is a drawing showing an early disease detection system based on high-depth cell-free nucleic acid sequencing according to an embodiment of the present invention. [Figure 2] Figure 2 is a drawing for explaining the physical configuration of an early disease detection apparatus based on high-depth cell-free nucleic acid sequencing according to an embodiment of the present invention. [Figure 3] Figure 3 is a drawing for explaining the functional configuration of an early disease detection apparatus based on high-depth cell-free nucleic acid sequencing according to an embodiment of the present invention. [Figure 4] Figure 4 is a drawing comparing high-depth full-length genome sequencing and general full-length genome sequencing. [Figure 5a] Figure 5a shows an example of detecting cancer by high-depth cell-free nucleic acid sequencing based on an embodiment of the present invention. [Figure 5b] Figure 5b shows an example of detecting cancer by high-depth cell-free nucleic acid sequencing based on an embodiment of the present invention. [Figure 6a] Figure 6a is a drawing showing a method of classifying cancer patients based on cancer-specific signal values according to an embodiment of the present invention. [Figure 6b] Figure 6b is a drawing showing a method of classifying cancer patients based on cancer-specific signal values according to an embodiment of the present invention. [Figure 6c] Figure 6c is a diagram illustrating a method for classifying cancer patients based on cancer-specific signal values, according to one embodiment of the present invention. [Figure 7] Figure 7 shows a specific example of disease-specific signal values ​​for generating disease prediction signal values ​​according to one embodiment of the present invention. [Figure 8a] Figure 8a is a diagram illustrating disease prediction signal values ​​according to one embodiment of the present invention. [Figure 8b] Figure 8b is a diagram illustrating disease prediction signal values ​​according to one embodiment of the present invention. [Figure 9a] Figure 9a is a diagram illustrating the procedure for an early disease detection method based on deep cell-free nucleic acid sequencing according to one embodiment of the present invention. [Figure 9b] Figure 9b is a diagram illustrating the procedure for an early disease detection method based on deep cell-free nucleic acid sequencing according to one embodiment of the present invention. [Modes for carrying out the invention]

[0027] Since the descriptions relating to the present invention are merely examples for structural or functional explanation, the scope of the present invention should not be interpreted as being limited by the examples described herein. That is, since the examples are subject to various modifications and can take on various forms, the scope of the present invention should be understood to include equivalents that can realize the technical idea. Furthermore, the purposes or effects shown in the present invention do not mean that a particular example must include all of them, or should only include such effects, so the scope of the present invention should not be interpreted as being limited by them.

[0028] The terms used in this application shall be understood as follows:

[0029] Terms such as "first" and "second" are used to distinguish one component from another, and the scope of rights should not be limited by these terms. For example, the first component may be named the second component, and similarly, the second component may be named the first component.

[0030] When it is stated that one component is "connected" to another component, it should be understood that this could mean that it is directly connected to the other component, or that another component exists in between. Conversely, when it is stated that one component is "directly connected" to another component, it should be understood that there is no other component in between. Other expressions describing the relationship between components, such as "between" and "directly between," or "adjacent to" and "directly adjacent to," should be interpreted similarly.

[0031] A singular expression shall be interpreted as including multiple expressions unless the context clearly indicates a different meaning, and terms such as "includes" or "possesses" shall be understood as indicating the existence of a feature, numerical value, stage, operation, component, part, or combination thereof, and shall not preemptively exclude the possibility of the existence or addition of one or more other features, numerical values, stages, operations, components, parts, or combinations thereof.

[0032] The identification codes (e.g., a, b, c, etc.) at each stage are assigned for explanatory purposes only and do not indicate the order of the stages. Unless a specific order is explicitly stated in the context, the stages may occur in a different order than that described. That is, the stages may occur in the same order as described, may be carried out substantially simultaneously, or may be carried out in the reverse order.

[0033] The present invention can be embodied in a computer-readable recording medium as a computer-readable code, and a computer-readable recording medium includes all kinds of recording devices on which data readable by a computer system is stored. Examples of computer-readable recording media include ROM (Read Only Memory), RAM (Random Access Memory), CD-ROM, magnetic tape, floppy disk, optical data storage devices, etc. Furthermore, computer-readable recording media can be distributed across computer systems connected to a network, and computer-readable code can be stored and executed in a distributed manner.

[0034] All terms used herein have the same meaning as commonly understood by experts in the art to which this invention pertains, unless otherwise defined. Terms as defined in dictionaries commonly used herein should be interpreted as having the same meaning as in the context of the relevant art, and should not be interpreted as having an ideal or overly formal meaning unless explicitly defined herein.

[0035] Figure 1 is a diagram showing a disease detection system 100 based on deep cell-free nucleic acid sequencing according to one embodiment of the present invention.

[0036] Referring to Figure 1, the disease detection system 100 based on deep cell-free nucleic acid sequencing may include a user terminal 110, a disease detection device 130 based on deep cell-free nucleic acid sequencing, and a database 150.

[0037] The user terminal 110 can be implemented as a smartphone or wearable device that can view data generated based on the generation operation of an early disease detection device, which creates a library set based on cell-free nucleic acids via a disease detection device 130 based on deep cell-free nucleic acid sequencing, generates deep cell-free nucleic acid sequencing data for the library set, and generates disease-specific signal values ​​and disease prediction signal values ​​based on the mapped deep cell-free nucleic acid sequencing data to generate a disease prediction model that predicts the presence or absence of disease and the type of disease in the blood of the person being tested, as well as metadata analyzed therefrom. It is not limited to these, and can also be implemented as a variety of devices such as tablet PCs. The user terminal 110 can be connected to the disease detection device 130 based on deep cell-free nucleic acid sequencing via a network, and multiple user terminals 110 can be connected to the disease detection device 130 simultaneously.

[0038] The disease detection device 130 based on deep cell-release nucleic acid sequencing can be implemented as a server equivalent to a computer or program that sequentially executes operations to generate an early disease detection device. The disease detection device 130 based on deep cell-release nucleic acid sequencing is wirelessly connected to the user terminal 110 via Bluetooth®, Wi-Fi®, a communication network, etc., and can exchange data with the user terminal 110 via the network.

[0039] The disease detection device 130 based on deep cell-free nucleic acid sequencing may be provided by recording a program consisting of instructions for implementing it on a computer-readable recording medium, thereby tangibly embodying the program. In other words, the disease detection device 130 based on deep cell-free nucleic acid sequencing may be embodied in the form of program instructions that can be executed by various computer means and recorded on a computer-readable recording medium. Furthermore, the disease detection device 130 based on deep cell-free nucleic acid sequencing consists of a computer program that sequentially executes operations for early detection of disease, and this computer program may be stored on a computer-readable recording medium.

[0040] Database 150 is a storage device that stores various information generated by the generation operation of the disease early detection device.

[0041] Figure 2 is a diagram showing the physical configuration of a disease detection device 130 based on deep cell-free nucleic acid sequencing according to one embodiment.

[0042] Referring to Figure 2, the disease detection device 130 based on deep cell-free nucleic acid sequencing may be configured to include a processor 210, memory 230, user input / output unit 250, and network input / output unit 270.

[0043] The processor 210 executes procedures to perform operations that generate an early disease detection device, manages the memory 230 which is read from and written throughout the process, and can schedule synchronization times between volatile and non-volatile memory in the memory 230. The processor 210 controls the overall operation of the disease detection device 130 based on deep cell-free nucleic acid sequencing, is electrically connected to the memory 230, the user input / output unit 250, and the network input / output unit 270, and can control the data flow between them. The processor 210 is embodied in the CPU (Central Processing Unit) or GPU (Graphic Processing Unit) of the disease detection device 130 based on deep cell-free nucleic acid sequencing.

[0044] The memory 230 is implemented as a non-volatile memory such as an SSD (Solid State Drive) or HDD (Hard Disk Drive) and may include an auxiliary storage device that stores all the data necessary for analysis based on deep cell free nucleic acid sequencing by the disease detection device 130, and may also include a main memory device implemented as a volatile memory such as RAM.

[0045] The user input / output unit 250 may include an environment for receiving user input and an environment for outputting specific information to the user. For example, the user input / output unit 250 may include an input device including an adapter such as a touchpad, touchscreen, virtual keyboard, or pointing device, and an output device including an adapter such as a monitor or touchscreen. In one embodiment, the user input / output unit 250 may correspond to a computing device connected via a remote connection, in which case the disease detection device 130 based on deep cell-free nucleic acid sequencing may function as a server.

[0046] The network input / output unit 270 includes an environment for connecting to external devices or systems via a network, and may include adapters for communication such as LAN (Local Area Network), MAN (Metropolitan Area Network), WAN (Wide Area Network), and VAN (Value Added Network).

[0047] Figure 3 is a diagram illustrating the functional configuration of a disease detection device 130 based on deep cell-free nucleic acid sequencing according to one embodiment of the present invention. Referring to Figure 3, the disease detection device 130 based on deep cell-free nucleic acid sequencing may include a library production unit 310, a sequencing unit 320, a signal calculation unit 330, a predicted signal value generation unit 340, a predicted model generation unit 350, and a disease onset prediction unit 360.

[0048] The library production unit 310 can produce library sets including whole genome sequencing libraries, whole epigenome sequencing libraries, transcriptome sequencing libraries, and epitranscriptome sequencing libraries based on cell-free nucleic acids extracted from blood.

[0049] Here, cell-free nucleic acids are extracted from plasma separated from blood. Cell-free nucleic acids may include cell-free DNA and cell-free RNA, and the cell-free DNA and cell-free RNA may include exosome DNA and exosome RNA, respectively. For example, cell-free nucleic acids are extracted using a Circulating Nucleic Acid Kit.

[0050] Blood can be extracted separately from individuals without disease, individuals with a history of a specific disease but who are currently free of that disease, patients in the early, middle, and late stages of a specific disease, and patients with a relapse of a specific disease. This is because, in addition to distinguishing between the presence or absence of disease, the characteristics of the blood differ depending on the stage of disease progression and whether or not treatment has been administered.

[0051] The library production unit 310 can produce a library set for cell-free nucleic acids extracted from blood using a library production kit, through an adapter ligation step and a library quantification and size confirmation step. Furthermore, the library production unit 310 can generate a whole genome bisulfite sequencing library as one method for producing a full-length epigenetic sequencing library based on cell-free nucleic acids. For example, the library production unit 310 can produce EpiNext TMHigh-Sensitivity DNA Library Preparation Kit, EZ DNA Methylation-Lightning MagPrep Kit, or EpiNext TM Library sets can be prepared using the Post-Bisulfite DNA Library Preparation Kit. Furthermore, the library preparation unit 310 can generate transcriptome bisulfite sequencing libraries as one method for preparing full-length epigenetic transcriptome sequencing libraries based on cell-free nucleic acids. For example, the library preparation unit 310 can use AMPINEXT TM 5-mC RNA Bisulfite-Seq Easy Kit or EpiNext TM A library set can be prepared using the RNA Bisulfite-Seq Kit.

[0052] Figure 4 is a diagram comparing deep-depth full-length genome sequencing with general full-length genome sequencing.

[0053] Referring to Figure 4, the sequencing unit 320 can sequence a library set to generate high-depth genome sequencing data. The sequencing unit 320 can generate genome sequencing data at a higher depth than general genome sequencing for the entire genome. Compared to panel sequencing, which sequences only specific regions of the genome, the sequencing unit 320 can sequence the entire genome at a higher depth and generate sequencing data. As a result, the disease detection device 130 based on high-depth cell-free nucleic acid sequencing according to the present invention can detect diseases with high accuracy based on high-depth sequencing data for the entire genome.

[0054] In one embodiment, the sequencing unit 320 can sequence the library set at a depth of 10 or more. Here, depth refers to the average number of sequencing reads that have specific location information on the genome. As the depth increases, the number of sequencing reads that have specific location information on the genome increases, resulting in a higher quantity and accuracy of the genomic information obtained.

[0055] Figure 5 shows the results of predicting cancer after extracting cancer-specific signals by full-length genome sequencing of free nucleic acids in the blood at various sequencing depths. The present invention can be applied not only to cancer, as in the example in Figure 5, but also to other diseases.

[0056] Figure 5a shows the prediction results when cancer prediction signal values ​​are generated using DNA signal values.

[0057] The dataset used for training consisted of data from 18 cancer patients and 18 healthy individuals. The dataset used to evaluate the training results consisted of data from 8 cancer patients and 8 healthy individuals. The predicted cancer was colorectal cancer.

[0058] Experimental results show that, according to the present invention, when full-length genome sequencing data with a sequencing depth of 10 or greater, i.e., sequencing data at depths of 10, 30, and 50, is used for cell-free nucleic acids in the blood, the cancer prediction signal scores of cancer patients and healthy individuals are more clearly distinguished. On the other hand, when full-length genome sequencing data with a low depth of less than 10 is used, the boundary between cancer prediction signal scores of cancer patients and healthy individuals is unclear.

[0059] Figure 5b shows the prediction results when cancer prediction signal values ​​are generated using DNA deformation signal values ​​in addition to DNA signal values.

[0060] When tested on colorectal cancer under the same conditions as in Figure 5a, it was confirmed that, similar to when only DNA signal values ​​were used, full-length genome sequencing of cell-free nucleic acids in the blood at a sequencing depth of 10 or more resulted in a clearer distinction between cancer prediction scores for cancer patients and healthy individuals.

[0061] In one embodiment, the sequencing unit 320 can utilize one or more sequencers to sequence the library set. Examples of available sequencers include Illumina, MGI, Element Biosciences, Ultima Genomics, PacBio, Oxford Nanopore, etc. Furthermore, various sequencers can be used without being limited by the length of the sequencing reads or sequencing techniques (e.g., chain termination, ligation, fluorescence, single / paired-end, etc.).

[0062] The signal calculation unit 330 can map high-depth cell-free nucleic acid sequencing data to a genome reference to determine the genomic location of the sequencing reads. More specifically, the signal calculation unit 330 can map the generated high-depth cell-free nucleic acid sequencing data to an appropriate genome reference so that its genomic location can be identified and analyzed. In other words, the signal calculation unit 330 can identify the location of the generated high-depth cell-free nucleic acid sequencing data so that it can be analyzed.

[0063] In one embodiment, the signal calculation unit 330 can map deep cell-free nucleic acid sequencing data to a reference genome. The deep cell-free nucleic acid sequencing data mapped to the reference genome is saved as a file such as a SAM (Sequence Alignment Map), BAM (Binary Alignment Map), or CRAM (Compressed Reference-oriented Alignment Map). The reference genome to which deep cell-free nucleic acid sequencing data is mapped is not limited to the human genome reference. To detect infectious diseases, the data can be mapped to a non-human genome reference (virus, bacteria, fungi, etc.) to confirm the extent of the genome sequence presence of the source of infection. For example, deep cell-free nucleic acid sequencing data can be mapped to pneumococcal reference genome data to detect bacterial pneumonia.

[0064] The signal calculation unit 330 can calculate at least one disease-specific signal value based on mapped deep cell-free nucleic acid sequencing data. The signal calculation unit 330 can analyze the deep cell-free nucleic acid sequencing data to calculate disease-specific signal values ​​related to the occurrence of disease. More specifically, disease-specific signal values ​​may show similar patterns or different patterns depending on the type of disease.

[0065] Figure 6 shows a method for classifying subjects based on disease-specific signal values ​​according to one embodiment of the present invention.

[0066] For example, in Figure 6a, the DNA deformation signal value (using the degree of positional methylation change), which is one of the disease-specific signal values, may be associated with colorectal cancer, and healthy individuals and cancer patients can be distinguished based on a specific cancer detection threshold (cutoff).

[0067] For example, in Figure 6b, the DNA signal value (utilizing the mutation site and the surrounding base sequence), which is one of the cancer-specific signal values, is associated with pancreatic cancer, and healthy individuals and cancer patients can be distinguished based on specific cancer detection thresholds.

[0068] As another example, DNA signal values ​​(fragment length distribution), one of the cancer-specific signal values ​​shown in Figure 6c, may be associated with testicular cancer, liver cancer, bladder cancer, breast cancer, and prostate cancer, and healthy individuals and cancer patients can be distinguished based on specific cancer detection thresholds.

[0069] The signal calculation unit 330 can calculate disease-specific signal values ​​that can distinguish between healthy individuals and patients with a particular disease. As described later, it can combine one or more disease-specific signal values ​​to generate disease prediction signal values, and multiple diseases can be detected using these disease prediction signal values. This will be explained in detail below.

[0070] In one embodiment, the signal calculation unit 330 can calculate disease-specific signal values ​​based on DNA signal values, DNA modification signal values, RNA signal values, and RNA modification signal values.

[0071] Figure 7 shows an example of data used to calculate these disease-specific signal values.

[0072] The signal calculation unit 330 can calculate DNA signal values ​​using high-depth, full-length cell free nucleic acid sequencing data sequenced from a full-length genome sequencing library.

[0073] The DNA signal value may include one or more of the following: the amount of free DNA extracted from blood, the mutation site and the set of base sequences surrounding it, gene copy number variation, disease-specific mutations, the genome sequence of the detected pathogen, the distribution of fragment lengths, sequence information around the fragment ends, the distribution of individual disease-specific nucleosomes obtained from the positional information of the fragment ends, and the jagged end ratio.

[0074] More specifically, the signal calculation unit 330 filters out existing single nucleotide polymorphism (SNP) locations and filters out low-quality mutation locations by considering factors such as base quality, mapping quality, and depth on the sequencing reads. The set of base sequences of the filtered mutation locations, the set of base sequences surrounding the mutation location (for example, 2X+1 consecutive bases from -Xnt to +Xnt at the mutation location), the set of sequencing read information for the mutation location and its surroundings, the correlation coefficient level when comparing the set of base sequences of the mutation location with a well-known disease-specific mutation reference (for example, the mutational signature of COSMIC for cancer), and the presence or absence of disease-specific mutations can be used to calculate the DNA signal value. Existing studies have observed significant differences between mutation locations and the surrounding base sequence sets between patients and non-patients in diseases such as cancer and clonal hematopoiesis (Bruhm et al., 2023, Nat. Genet.; Fairchild et al., 2023, Sci. Transl. Med.). These mutations include structural variations consisting of single base substitutions, doublet base substitutions, small and large insertions and deletions, tandem duplication, inversion, fusion, translocation, chromoplexy, and chromothripsis. In the case of structural variations, the breakpoint location, which is a discontinuous point on the genome in a sequencing read, and the surrounding base sequence set can be used as DNA signal values.

[0075] The signal calculation unit 330 can utilize pathogen genome sequences detected from high-depth full-length cell free nucleic acid sequencing data to calculate DNA signal values. Pathogens may include pathogenic DNA viruses, bacteria, and fungi that can cause infectious diseases.

[0076] The signal calculation unit 330 can utilize the degree of gene copy number alteration and the degree of chromosomal aneuploidy detected from cell-free nucleic acid sequencing data of the full-length, deep genome to calculate DNA signal values.

[0077] The signal calculation unit 330 can calculate DNA signal values ​​using fragment information obtained from high-depth cell-free nucleic acid sequencing data.

[0078] More specifically, the distribution of fragment lengths, the set of sequence information of Y consecutive bases around the fragment ends, and the disease-specific nucleosome distribution obtained from the positional information of the fragment ends on the genome can be used to generate DNA signal values. Existing studies have shown significant differences in fragment signals between patients and non-patients in diseases such as cancer, systemic lupus erythematosus, and fatty liver (Cristiano et al., 2019; Ding et al., 2022; Zhou et al., 2023; Karlas et al., 2017).

[0079] The signal calculation unit 330 can use the amount of free DNA extracted from blood in the library production unit 310 to calculate the DNA signal value.

[0080] More specifically, the amount of free DNA in cells can be determined using methods such as electrophoresis and fluorescent labeling. For example, the amount of free DNA can be determined using an Agilent 2100 Bioanalyzer or an Invitrogen Qubit 4.0 fluorometer. Existing studies have shown significant differences in the amount of free DNA in cells between patients and non-patients in diseases such as cancer, systemic lupus erythematosus, rheumatoid arthritis, hypertension, and heart disease (Chen et al., 2021, Sci. Rep.; Truszewska et al., 2020, Lupus; Hashimoto et al., 2016, Int. J. Rheum.; Jeong et al., 2015, Kidney. Res. Clin. Pract.; Antonatos et al., 2006, Ann. NY Acad. Sci.).

[0081] The DNA deformation signal value may include at least one of the positional modification degrees of DNA chemical modifications, including 1mA, 1mHyp, 3mA, 3mHyp, 4mA, 6mA, 7mA, 7mHyp, 9mA, 9mHyp, 6hmA, m2A, ncm6A, xI, m2C, e3C, 3mC, εC, 4mC, 5caC, 5fC, 5hmC, 5mC, 1mG, 2mG, 3mG, m6G, 7mG, CEG, M22G, N(2),3-εG, 1,N(2)-εG, ADG, 8oxoG, dU, dhpdU, 5-NedU, DHdU, 5-NeOmdU, 5caU, dhpUra, 5hmU, 5fU, base J, diHT, 1mT, 3mT, 4-meT, and putThy.

[0082] The signal calculation unit 330 can calculate DNA deformation signal values ​​using deep-depth full-length epigenetic sequencing data obtained by sequencing a sequencing library specifically designed to detect specific chemical deformations of DNA (including, for example, 5mC, 5hmC, 6mA, etc.), or deep-depth full-length genome sequencing data based on a specific sequencer (e.g., a sequencer from Oxford Nanopore) that can directly read DNA deformation information from a full-length genome sequencing library.

[0083] For example, to detect 5mC, a type of DNA methylation modification, a full-length genome bisulfite sequencing library can be constructed and used to obtain deep cell-free nucleic acid sequencing data sequenced with an Illumina sequencer, or deep cell-free nucleic acid sequencing data can be obtained by sequencing a full-length genome sequencing library with an Oxford Nanopore sequencer.

[0084] More specifically, the signal calculation unit 330 obtains DNA chemical modification locations and locations where DNA chemical modification does not occur from high-depth full-length epigenetic sequencing data or a full-length genome sequencing library based on a specific sequencer, and filters out existing known single nucleotide polymorphism locations or removes inappropriate DNA modification locations by considering the base quality, mapping quality, and depth on the sequencing reads. Subsequently, by comparing the DNA chemical modification signals of diseased patients and healthy individuals, it selects DNA chemical modification regions (differentially modified regions, DMRs) that differ significantly in diseased patients, and calculates the degree of DNA chemical modification at the relevant locations as a DNA deformation signal value. The signal calculation unit 330 can then reflect the degree of DNA chemical modification at the relevant locations, sequencing read information, etc., in the DNA deformation signal value. Existing studies have shown significant differences in methylation signals between patients and non-patients in diseases such as cancer, systemic lupus erythematosus, rheumatoid arthritis, hypertension, heart disease, and fatty liver (Shen et al., 2018, Nature; Wang et al., 2022, Front. Immunol.; Plant et al., 2016, Arthritis Rheumatol.; Brusca et al., 2022, Circulation; Zemmour et al., 2018, Nat. Commun.; Yigit et al., 2018, Gut).

[0085] The signal calculation unit 330 can calculate RNA signal values ​​using high-depth transcriptome sequencing data obtained from a transcriptome sequencing library.

[0086] The RNA signal value includes at least one of the following: the degree of expression of individual disease-specific genes, the degree of expression of fusion genes, and the degree of alternative splicing.

[0087] More specifically, the signal calculation unit 330 can remove low-quality sequencing reads by considering the base quality, mapping quality, and depth on the sequencing reads, and calculate the RNA signal value after confirming whether there are enough sequencing reads that can support the expression of genes known to induce disease. In addition, the signal calculation unit 330 can utilize the gene expression level of disease-specific genes, the expression level of fusion genes, the type of fusion gene being expressed, and the degree of alternative splicing to calculate the RNA signal value. Alternative splicing can include one or more of the following: intron retention, exon skipping, alternative 5' splicing, alternative 3' splicing, mutually exclusive splicing, alternative transcription initiation, and alternative polyadenylation. Existing studies have shown significant differences in the expression levels of disease-specific genes and fusion genes, as well as the degree of alternative splicing, between patients and non-patients in diseases such as cancer (Larson et al., 2021, Nat. Commun.; Hasegawa et al., 2021, Cancer Sci.; Zhu et al., 2021, Theranostics).

[0088] The RNA modification signal value may include any one or more of the modification degrees at each position of RNA chemical modifications including ms2m6A, io6A, m6t6A, hn6A, Ar(p), m1I, m1Am, m1Im, m2A, m62A, ms2i6A, ms2t6A, i6A, ms2hn6A, m6Am, m1A, m6A, I, g6A, m62Am, ac6A, ms2io6A, Am, Im, t6A, m8A, m2,8A, ct6A, hm6A, f6A, ms2ct6A, ht6A, msms2i6A, m5Cm, m4C, m42C, m5C, f5C, m4Cm, m3C, m42Cm, ac4C, Cm, s2C, k2C, f5Cm, ac4Cm, hm5C, C+, ho5C, hm5Cm, m2, 7G, oQ, m22G, m1G, Gm, yW, Gr(p), m2G, m22Gm, gluQ, mimG, preQ1base, m1Gm, Q, yW-86, yW-72, galQ, imG2, manQ, OHyW, G+, m2Gm, imG-14, m227G, yW-58, m7G, preQ0, o2yW, preQ1, imG, m2, 7Gm, f5U, m5s2U, s4U, m1acp3Y, cm5s2U, m5U, D, m5D, tm5U, cmnm5Um, f5Um, mcm5U, ncm5Um, nm5s2U, ncm5U, cm5U, cmnm5s2U, cmo5U, m3Y, m3U, Um, acp3Um, cmnm5se2U, m3Um, mcmo5U, nm5se2U, ho5U, f5s2U, mcm5s2U, chm5U, mo5U, acp3U, mnm5se2U, mcm5Um, mchm5U, cmnm5U, m1Y, tm5s2U, se2U, s2U, Ym, f5se2U, mnm5s2U, nm5U, s2Um, Y, m5Um, mnm5U, acp3D, acp3Y, inm5U, inm5Um, inm5s2U, ncm5s2U, nchm5U, mchm5Um, ges2U, cmnm5ges2U, mnm5ges2U, nm5ges2U, cnm5U, mcmo5Um.

[0089] The signal calculation unit 330 performs specific chemical modifications of RNA (e.g., m 6 A, m 5 C, Ψ, I, ac 4 C, Nm, m 6 Am, m 1A,m 7 G, hm 5 RNA deformation signal values ​​can be calculated using deep-depth epigenetic transcriptome sequencing data obtained by sequencing a sequencing library specifically designed to detect (e.g., C), or deep-depth transcriptome sequencing data based on a specific sequencer (e.g., a sequencer from Oxford Nanopore) that can directly read RNA chemical deformation information from a transcriptome genome sequencing library.

[0090] For example, m 5 For 14C detection, a transcriptome bisulfite sequencing library can be fabricated and used to obtain deep cell-free nucleic acid sequencing data sequenced with an Illumina sequencer, or a full-length genome sequencing library can be used to obtain deep cell-free nucleic acid sequencing data sequenced with an Oxford Nanopore sequencer.

[0091] More specifically, the signal calculation unit 330 obtains RNA chemical modification sites and sites without RNA chemical modification from high-depth epitranscriptome sequencing data or a transcriptome sequencing library based on a specific sequencer, filters out known single nucleotide polymorphism (SNP) locations, or removes inappropriate RNA chemical modification sites considering the base quality, mapping quality, and sequencing depth on the sequencing read. Subsequently, it compares the RNA chemical modification signals of diseased patients and healthy individuals to select RNA chemical modification regions that differ significantly in diseased patients, and calculates the degree of RNA chemical modification at those locations as an RNA signal value. Furthermore, the signal calculation unit 330 can reflect the degree of RNA chemical modification at those locations and sequencing read information in the RNA signal value.

[0092] Figure 8 is a diagram illustrating disease prediction signal values ​​according to one embodiment of the present invention.

[0093] Referring to Figure 8, the prediction signal value generation unit 340 can generate disease prediction signal values ​​by processing at least one disease-specific signal value. The prediction signal value generation unit 340 can generate disease prediction signal values ​​by standardizing disease-specific signal values, performing one-hot encoding, filtering low-quality data, and processing them by randomly selecting a portion or selecting a portion based on specific criteria. For example, the prediction signal value generation unit 340 can generate disease prediction signal values ​​according to the type of target disease by processing DNA signal values ​​and DNA deformation signal values.

[0094] In one embodiment, the prediction signal value generation unit 340 can generate a disease prediction signal value by processing multiple disease prediction signal values ​​into vector data. For example, the prediction signal value generation unit 340 can represent multiple disease prediction signal values ​​a, b, c, and d as vector values ​​by specifying their respective positions. More specifically, the prediction signal value generation unit 340 can generate a disease prediction signal value by selecting and processing specific disease-specific signal values ​​depending on the type of target disease.

[0095] In one embodiment, the prediction signal value generation unit 340 can generate disease prediction signal values ​​according to the type of target disease. As described above, the prediction signal value generation unit 340 can select disease-specific signal values ​​to be used according to the type of target disease, process them, and generate disease prediction signal values.

[0096] In the example shown in Figure 8a, the prediction signal value generation unit 340 can generate a disease prediction signal value by combining the DNA signal value, DNA deformation signal value, RNA signal value, and RNA deformation signal value for disease A.

[0097] In the example shown in Figure 8b, the prediction signal value generation unit 340 can generate a cancer prediction signal value by combining the DNA signal value and DNA deformation signal value calculated for colorectal cancer, and can also generate a cancer prediction signal value by combining the DNA deformation signal value and RNA signal value calculated for pancreatic cancer.

[0098] More specifically, the disease-specific signal values ​​calculated differ depending on the type of target disease. As a result, the predictive signal value generation unit 360 can process the disease-specific signal values ​​to generate disease prediction signal values ​​according to the type of target disease. Therefore, the predictive signal value generation unit 340 can detect various types of diseases through various combinations of DNA signal values, DNA deformation signal values, RNA signal values, and RNA deformation signal values.

[0099] The prediction model generation unit 350 can generate a disease prediction model that detects whether or not a disease has occurred in the input blood sample, using labeling information and disease prediction signal values ​​related to the blood.

[0100] More specifically, the prediction model generation unit 350 can create a disease prediction model by learning using labeling information related to blood and disease prediction signal values ​​generated via the blood. Here, "created by learning" means that the artificial intelligence model is trained using a large amount of training data by a learning algorithm, and as a result, a predefined set of behavioral rules or artificial intelligence model is created that is configured to perform a desired characteristic (or objective). Such learning may be performed on the device on which the artificial intelligence is executed based on this disclosure, or it may be performed via a separate server and / or system. Examples of learning algorithms include, but are not limited to, supervised learning, unsupervised learning, semi-supervised learning, or reinforcement learning.

[0101] An artificial intelligence model can consist of multiple neural network layers. Each layer of the neural network has multiple weights, and neural network operations are performed by calculating the results of operations in previous layers and these weights. The multiple weights of the multiple neural network layers can be optimized as a result of the AI ​​model's learning. For example, the multiple weights can be updated so that the loss or cost value obtained by the AI ​​model during the learning process is reduced or minimized.

[0102] Artificial neural networks can include deep neural networks (DNNs), and are not limited to, but include, for example, CNNs (Convolutional Neural Networks), DNNs (Deep Neural Networks), RNNs (Recurrent Neural Networks), RBMs (Restricted Boltzmann Machines), DBNs (Deep Belief Networks), BRDNNs (Bidirectional Recurrent Deep Neural Networks), or DQNs (Deep Q-Networks). They can also include learning algorithms such as linear regression, logistic regression, probit regression, naive Bayes, Bayes networks, partial least squares, k-nearest neighbors, genetic programming, decision trees, random forests, and support vector machines.

[0103] Furthermore, the artificial intelligence models used in the disclosed embodiments can be implemented in various embodiments by manufacturers or users of electronic devices, and are not limited to the examples given above.

[0104] The disease onset prediction unit 360 can input blood data extracted from the person being tested and detect whether or not the person has developed a disease using a disease detection model. For example, the disease detection model generated by the prediction signal value generation unit 340 can be input with high-depth cell-free nucleic acid sequencing data generated based on cell-free nucleic acids extracted from the blood of a specific person being tested to detect whether or not the person has developed a disease.

[0105] On the other hand, the blood of the person being tested, whose presence or absence of disease is detected by the disease onset prediction unit 360, must not be used in the disease detection model that has already been generated via the prediction signal value generation unit 340. This is to prevent overfitting of the disease detection model.

[0106] For example, a library set can be prepared from cell-free nucleic acids extracted from the blood of the person being tested via the library preparation unit 310, the library set can be sequenced via the sequencing unit 320, the high-depth cell-free nucleic acid data can be mapped to a genome reference via the signal calculation unit 330, and disease prediction signal values ​​can be generated from the mapped sequencing data via the prediction signal value generation unit 340. Furthermore, the disease prediction signal values ​​generated by the disease onset prediction unit 360 and the already generated disease prediction model can be used to predict whether or not the person being tested will develop the disease and the type of disease.

[0107] Figure 9 shows the procedure for a disease detection method based on deep cell-free nucleic acid sequencing according to one embodiment of the present invention.

[0108] Figure 9A shows a schematic flowchart of a method for generating disease prediction models based on deep cell-free nucleic acid sequencing.

[0109] Referring to 9A, the disease detection method based on deep cell-free nucleic acid sequencing can produce a library set including a full-length genome sequencing library, a full-length epigenome sequencing library, and a transcriptome sequencing library based on cell-free nucleic acids extracted from blood via the library production unit 310 (S110).

[0110] The disease detection method based on deep cell-free nucleic acid sequencing can generate deep cell-free nucleic acid sequencing data by sequencing a library set via the sequencing unit 320 (S120).

[0111] The disease detection method based on deep cell-free nucleic acid sequencing can use a signal calculation unit 330 to map deep cell-free nucleic acid sequencing data to a genome reference to identify the genomic location of the data, and to calculate at least one disease-specific signal value based on the mapped data (S130).

[0112] The disease detection method based on deep cell-free nucleic acid sequencing can generate a disease prediction signal value by processing at least one disease-specific signal value via the prediction signal value generation unit 340 (S140).

[0113] Referring to Figure 9A, the disease detection method based on deep cell-free nucleic acid sequencing can generate a disease prediction model via the prediction model generation unit 350, which uses labeling information related to blood and disease prediction signal values ​​to detect the presence or absence of disease and the type of disease in the input blood (S150).

[0114] Figure 9b is a schematic flowchart of a method for predicting disease using a disease prediction model generated based on deep cell-free nucleic acid sequencing.

[0115] Referring to Figure 9b, the disease detection method based on deep cell-free nucleic acid sequencing can detect the presence or absence of disease and the type of disease in the input blood via the disease prediction unit 360, using labeling information related to the blood, disease prediction signal values, and a pre-generated disease prediction model (S250).

[0116] While preferred embodiments of the present invention have been described above with reference to the present invention, experts skilled in the art will understand that the present invention can be modified and altered in various ways without departing from the spirit and scope of the invention as set forth in the following claims. [Explanation of symbols]

[0117] 100: Disease detection system based on deep cell-free nucleic acid sequencing 110: User terminal 130: Disease detection device based on deep cell-free nucleic acid sequencing 150: Database 210: Processor 230: Memory 250: User Input / Output Section 270: Network Input / Output Section 310: Library Production Department 320: Sequencing Department 330: Signal Calculation Unit 340: Predicted signal value generation unit 350: Predictive Model Generation Unit 360: Disease Prediction Department

Claims

1. A library production department that prepares library sets including a full-length genome sequencing library, a full-length genome epigenetic sequencing library, and a transcriptome sequencing library based on cell-free nucleic acids extracted from blood, A sequencing unit that sequences the aforementioned library set at a high depth of sequencing depth of 10 or more to generate high-depth cell-free nucleic acid sequencing data, A signal calculation unit maps the aforementioned deep cell-free nucleic acid sequencing data to a genome reference to identify the location of the data on the genome, and calculates at least one disease-specific signal value based on the mapped data. A prediction signal value generation unit that processes the at least one disease-specific signal value to generate a disease prediction signal value, The system includes a prediction model generation unit that generates a disease prediction model for detecting the presence and type of disease in the input blood using the labeling information related to the blood and the disease prediction signal value. A disease early detection device based on deep cell-free nucleic acid sequencing.

2. The disease early detection device based on deep cell-free nucleic acid sequencing according to claim 1, characterized in that the blood is extracted separately from persons without disease, persons with a history of a specific disease but who are currently free from that disease, patients in the early, middle, and late stages of a specific disease, and patients with a relapse of a specific disease.

3. The signal calculation unit, The disease-specific signal value is calculated based on one or more of the following: DNA signal value, DNA deformation signal value, RNA signal value, and RNA deformation signal value. A disease early detection device based on deep cell-free nucleic acid sequencing as described in claim 1.

4. The predicted signal value generation unit is: A disease early detection device based on deep cell-free nucleic acid sequencing according to claim 1, characterized in that it processes multiple disease-specific signal values ​​as vector data to generate disease prediction signal values.

5. The predicted signal value generation unit is: A disease early detection device based on deep cell-free nucleic acid sequencing according to claim 1, characterized in that it generates disease prediction signal values ​​according to the type of target disease.

6. The early disease detection device based on deep cell-free nucleic acid sequencing according to claim 1, further comprising a disease prediction unit that receives blood data extracted from a person under examination as input and detects whether or not the person under examination has developed a disease and the type of disease through the disease prediction model.

7. The disease early detection device based on deep cell-free nucleic acid sequencing according to claim 3, characterized in that the DNA signal value includes one or more of the following: the amount of cell-free DNA extracted from blood, the mutation site and the set of base sequences around it, gene copy number variation, chromosomal aneuploidy, disease-specific mutation, the genome sequence of the detected pathogen, the distribution of fragment lengths, sequence information around the fragment ends, the distribution of individual disease-specific nucleosomes obtained from the positional information of the fragment ends, and the proportion of jagged ends.

8. The DNA deformation signal values ​​are 1mA, 1mHyp, 3mA, 3mHyp, 4mA, 6mA, 7mA, 7mHyp, 9mA, 9mHyp, 6hmA, m2A, ncm6A, xI, m2C, e3C, 3mC, εC, 4mC, 5caC, 5fC, 5hmC, 5mC. , 1mG, 2mG, 3mG, m6G, 7mG, CEG, M22G, N(2),3-εG, 1,N(2)-εG, ADG, 8oxoG, dU, dhpdU, 5-NedU, DHdU, 5-NeOmdU, 5caU, dhpUra, 5hmU, 5fU, base A disease early detection device based on deep cell-free nucleic acid sequencing according to claim 3, characterized in that it includes at least one of the positional modification degrees of chemical modifications of DNA, including J, diHT, 1mT, 3mT, 4-meT, and putThy.

9. The disease early detection device based on deep cell-free nucleic acid sequencing according to claim 3, characterized in that the RNA signal value includes at least one of the expression levels of individual disease-specific genes, the expression levels of fusion genes, and the degree of alternative splicing.

10. The RNA modification signal values ​​are ms2m6A, io6A, m6t6A, hn6A, Ar(p), m1I, m1Am, m1Im, m2A, m62A, ms2i6A, ms2t6A, i6A, ms2hn6A, m6Am, m1A, m6A, I, g6A, m62Am, ac6A, ms2io6A, Am, Im, t 6A, m8A, m2,8A, ct6A, hm6A, f6A, ms2ct6A, ht6A, msms2i6A, m5Cm, m4C, m42C, m5C, f5C, m4Cm, m3C, m42Cm, ac4C, Cm, s2C, k2C, f5Cm, ac4Cm, hm5C, C+, ho5C, hm5Cm, m2, 7G, oQ, m22G, m1G, Gm, yW, Gr(p), m2G, m22Gm, gluQ, mimG, preQ1base, m1Gm, Q, yW-86, yW-72 , galQ, imG2, manQ, OHyW, G+, m2Gm, imG-14, m227G, yW-58, m7G, preQ0, o2yW, preQ1, imG, m2, 7Gm, f5U, m5s2U, s4U, m1acp3Y, cm5s2U, m5U, D, m5D, tm5U, cmnm5Um, f5Um, mcm5U, ncm5Um, nm5s2U, ncm5U, cm5U, cmnm5s2U, cmo5U, m3Y, m 3U, Um, acp3Um, cmnm5se2U, m3Um, mcmo5U, nm5se2U, ho5U, f5s2U, mcm5s2U, chm5U, mo5U, acp3U, mnm5se2U, mcm5Um, mchm5U, cmnm5U, m1Y, A disease early detection device based on deep cell-free nucleic acid sequencing according to claim 3, characterized in that it includes one or more positional modification degrees from among tm5s2U, se2U, s2U, Ym, f5se2U, mnm5s2U, nm5U, s2Um, Y, m5Um, mnm5U, acp3D, acp3Y, inm5U, inm5Um, inm5s2U, ncm5s2U, nchm5U, mchm5Um, ges2U, cmnm5ges2U, mnm5ges2U, nm5ges2U, cnm5U, and mcmo5Um.

11. The disease early detection device based on deep cell-free nucleic acid sequencing according to claim 1, characterized in that the sequencing unit utilizes a plurality of sequencers to sequence the library set.

12. To construct a library set including a whole-genome sequencing library, a full-length genome epigenetic sequencing library, a transcriptome sequencing library, and an epigenetic transcriptome sequencing library based on cell-free nucleic acids extracted from blood, The steps include: sequencing the aforementioned library set at high depth to generate high-depth cell-free nucleic acid sequencing data; The process involves mapping the aforementioned deep cell-free nucleic acid sequencing data to a genome reference to identify the genomic location of the deep cell-free nucleic acid sequencing data, and calculating at least one disease-specific signal value based on the mapped deep cell-free nucleic acid sequencing data. The steps include: processing at least one disease-specific signal value to generate a disease prediction signal value; A step of generating a disease prediction model using the labeling information related to the blood and the disease prediction signal value, A disease detection method based on deep cell-free nucleic acid sequencing, including [the specified method].

Images