A donor-reference-free multigenomic cfDNA tracing system and its application
The multi-genomic cfDNA tracing system without donor reference solves the problem of difficulty in obtaining donor genotype information in existing technologies, enabling accurate and non-invasive monitoring of lung transplant rejection and personalized treatment guidance.
Patent Information
- Application Number
- CN202611343254.2
- Authority / Receiving Office
- CN · China
- Patent Type
- Applications(China)
- Current Assignee / Owner
- Filing Date
- 2026-09-01
- Publication Date
- 2026-11-17
AI Technical Summary
Existing diagnostic methods for lung transplant rejection require prior acquisition of genotype information from the recipient, allo-HSCT donor, and lung graft donor. Furthermore, donor samples are difficult to obtain in clinical practice, making it difficult to interpret traditional dd-cfDNA test results.
A multi-genomic cfDNA source determination system without donor reference is provided. It uses an SNP comparison analysis module to screen informative SNP sites, and combines a lung-derived dd-cfDNA proportion calculation module and a result output module to achieve dd-cfDNA analysis without prior acquisition of donor genotype information.
This method enables the identification and quantification of lung-derived dd-cfDNA in lung transplant recipients after allo-HSCT, improving the accuracy and stability of detection results, providing a non-invasive dynamic monitoring method, guiding individualized immunosuppressive therapy, and improving patient prognosis.
Smart Images

Figure CN123120031A_ABST
Abstract
Description
Technical Field
[0001] This invention relates to the field of organ transplant rejection risk assessment technology, and in particular to a multi-genomic cfDNA tracing system without donor reference and its application. Background Technology
[0002] Lung transplantation is the only effective treatment for end-stage lung disease, but post-transplant rejection is one of the main reasons limiting long-term survival and graft function. Currently, the gold standard for diagnosing rejection is pathological diagnosis following a lung biopsy. However, lung biopsies carry potential risks such as pneumothorax and bleeding, limiting their clinical application. Therefore, reliable non-invasive monitoring methods are urgently needed.
[0003] Donor-derived cell-free DNA (dd-cfDNA) is a circulating DNA fragment released into the bloodstream after graft cell damage. dd-cfDNA can sensitively reflect damage to the transplanted lung and has become an important non-invasive screening tool in the field of transplant rejection.
[0004] With the development of hematopoietic stem cell transplantation technology, the number of patients with respiratory failure due to chronic graft-versus-host disease after allogeneic hematopoietic stem cell transplantation (allo-HSCT) is on the rise, and the number of patients receiving lung transplantation after allogeneic hematopoietic stem cell transplantation is also increasing year by year, with children and young adults under 40 years old accounting for more than 80% of these cases. However, in lung transplant recipients after allogeneic hematopoietic stem cell transplantation, circulating cell-free DNA (cfDNA) not only originates from the recipient's tissue and lung graft but is also affected by hematopoietic cells from the allo-HSCT donor. Traditional lung transplant dd-cfDNA testing is mainly based on the theory that plasma cfDNA mainly originates from the recipient, with donor-derived cfDNA accounting for only a small proportion. However, in the lung transplant population after allo-HSCT, the main cfDNA in the plasma comes from the allo-HSCT donor, making it difficult to directly interpret the results of traditional dd-cfDNA testing in clinical practice. Furthermore, existing genotype-dependent analysis methods typically require prior acquisition of genotype information from the recipient, allo-HSCT donor, and lung graft donor to analyze the origin of cfDNA. However, in clinical practice, donor samples are often difficult to obtain.
[0005] Therefore, establishing a dd-cfDNA analysis strategy that does not require prior acquisition of lung transplant donor genotype information to identify and quantify lung-derived dd-cfDNA in lung transplant recipients after allo-HSCT, and exploring its application potential in lung transplant rejection monitoring, is of great clinical significance. Summary of the Invention
[0006] The purpose of this invention is to provide a multi-genomic cfDNA tracing and determination system without donor reference and its application, in order to solve the problem that existing analysis methods require prior acquisition of the genotype information of the recipient, allo-HSCT donor, and lung graft donor, while donor samples are often difficult to obtain in clinical practice.
[0007] To achieve the above objectives, the present invention provides a multi-genomic cfDNA source determination system without donor reference, the system comprising an SNP alignment analysis module, a lung-derived dd-cfDNA proportion calculation module, and a result output module.
[0008] Preferably, the SNP alignment analysis module analyzes SNP sites in quality-controlled sequencing data of plasma cfDNA, urine cfDNA, and peripheral blood leukocyte gDNA to screen out informative SNP sites.
[0009] Preferably, when analyzing SNP sites, quality-controlled reads are aligned to the human reference genome. After alignment, duplicate reads are removed and low-quality reads are filtered out. Variation identification and allele extraction are performed within the target SNP region. The target SNP region is a known SNP site linked to heterologous cfDNA.
[0010] Preferably, the method for screening informative SNP sites is as follows: For each target SNP site, the total number of reads, the number of reference allele reads, and the number of alternative allele reads are extracted from plasma cfDNA, urine cfDNA, and peripheral blood leukocyte gDNA, respectively. Then, the frequency of variant alleles in urine cfDNA and peripheral blood leukocyte gDNA is calculated using the following formula, and the genotypes of the recipient and allo-HSCT donor at the SNP site are estimated based on the frequency of variant alleles. SNP sites with the same estimated genotypes are screened as common background candidate sites. If an additional low-frequency allele signal different from the above common background pattern is detected in plasma cfDNA at this site, then the SNP site is regarded as an informative SNP site. The frequency of variant alleles = Alt.dp / All.dp, where Alt.dp is the number of alternative allele reads and All.dp is the total number of reads.
[0011] Preferably, when the frequency of a variant allele at a certain SNP site in urinary cfDNA is less than 10% or greater than 90%, the recipient is determined to be homozygous at that site; when the frequency of a variant allele at a certain SNP site in peripheral blood leukocyte gDNA is less than 10% or greater than 90%, the allo-HSCT donor is determined to be homozygous at that site.
[0012] Preferably, the informative SNP site is an SNP site in which the VAF in urine cfDNA is less than 10% or greater than 90%, the VAF in peripheral blood leukocyte gDNA is less than 10% or greater than 90%, and the presumed genotypes of both are the same, and the plasma cfDNA shows an additional low-frequency allele signal that is different from the common background pattern of the recipient and allo-HSCT donor.
[0013] Preferably, the lung-derived dd-cfDNA proportion calculation module calculates the prior probability that the lung transplant donor is homozygous or heterozygous at each informative SNP locus based on the allele frequency of a specific population. Then, the prior probability is incorporated into a binomial mixture model, and the proportion of lung-derived dd-cfDNA is solved by maximizing the log-likelihood function through maximum likelihood estimation or quasi-Newton method.
[0014] Preferably, the quasi-Newton method for maximizing the log-likelihood function is defined as follows: , in, and For each informative site Total number of reads, number of second allele reads; and These represent the prior probabilities that the lung transplant donor is homozygous or heterozygous at this site; It is the binomial probability mass function.
[0015] Preferably, the result output module outputs the result based on the result of the lung-derived dd-cfDNA proportion calculation module. The output result is either low risk of rejection or high risk of rejection. Referring to existing literature (see https: / / pubmed.ncbi.nlm.nih.gov / 38053854 / ), the threshold for low risk / high risk of rejection is set. A threshold greater than or equal to 1.17% is considered high risk of rejection, and a threshold less than 1.17% is considered low risk of rejection.
[0016] Instruments that include the aforementioned multi-genome cfDNA tracing and determination system without donor references.
[0017] Therefore, the present invention provides a donor-reference-free multi-genomic cfDNA tracing and determination system and its application, the specific technical effects of which are as follows: (1) This invention provides a method that, for the first time, eliminates the need to obtain genotype information of the recipient, allo-HSCT donor, and lung graft donor beforehand. It enables the identification and quantification of lung-derived dd-cfDNA in lung transplant recipients after allo-HSCT through the "exclusion method" and mathematical modeling, and accurately predicts whether rejection will occur after lung transplantation in allogeneic hematopoietic stem cell transplant recipients. This method breaks through the bottleneck of obtaining donor genotype information, solves the failure problem of traditional dual-genome methods, and makes the system clinically valuable. (2) The system provided by this invention uses a binomial mixed model and the maximum likelihood estimation method to improve the accuracy and stability of lung-derived dd-cfDNA quantification through joint analysis of multiple informative SNP sites, and avoids the randomness of single site analysis. (3) It can achieve non-invasive dynamic monitoring. Compared with invasive lung biopsy (which carries risks such as pneumothorax and bleeding), the system provided by this invention can perform repeated sampling and dynamic monitoring to track the changing trend of dd-cfDNA levels, providing clinicians with richer and more continuous information. Through accurate dd-cfDNA quantification, it can guide doctors to make individualized adjustments to immunosuppressive therapy, avoid excessive or insufficient immunosuppression, and improve patient prognosis. (4) The donor-free multi-genome cfDNA tracing system provided by the present invention provides a new approach to precision medicine by combining the integrated analysis of multi-source biological samples with mathematical modeling, and provides a new technical paradigm for solving complex medical problems.
[0018] The technical solution of the present invention will be further described in detail below with reference to the accompanying drawings and embodiments. Attached Figure Description
[0019] To more clearly illustrate the technical solutions of the embodiments of the present invention, the drawings used in the description of the embodiments of the present invention will be briefly introduced below. Obviously, the drawings described below are only some embodiments of the present invention. For those skilled in the art, other drawings can be obtained based on these drawings without creative effort.
[0020] Figure 1 This is a schematic diagram of a donor-reference-free multi-genomic cfDNA tracing and determination system provided by the present invention; Figure 2 These are the detection results of acute rejection samples and stable samples in Embodiment Six of the present invention; where A is the method of the prior art; and B is the method provided by the present invention. Figure 3This is a comparison of the identification performance of lung-derived dd-cfDNA before and after correction in Example 7 of the present invention; where A is the receiver operating characteristic curve; B is the precision-recall curve; the dashed line represents the result using the prior art, and the solid line represents the result using the method of the present invention; Figure 4 This is the correlation analysis result between the quantitative value of lung-derived dd-cfDNA obtained by the method of the present invention in Example 9 of the present invention and the quantitative value of lung-derived dd-cfDNA obtained based on lung tissue genotype reference. Detailed Implementation
[0021] The technical solution of the present invention will be further described below with reference to the accompanying drawings and embodiments.
[0022] To make the objectives, technical solutions, and advantages of this application clearer, more thorough, and more complete, the technical solutions of the present invention will be clearly and completely described below with reference to the accompanying drawings and embodiments. The following detailed descriptions are all illustrations of embodiments, intended to provide further detailed explanation of the present invention. Unless otherwise specified, all technical terms used in this invention have the same meaning as commonly understood by one of ordinary skill in the art to which this application pertains.
[0023] The instruments, equipment, reagents and materials used in the embodiments were all obtained through commercial means; the methods and steps not described in detail in the embodiments are all conventional techniques in the art.
[0024] The flowchart of a donor-reference-free multigenomic cfDNA tracing and determination system provided by this invention is as follows: Figure 1 As shown, it includes: Step 1: Sample collection and processing.
[0025] When patients undergoing allo-HSCT lung transplantation are undergoing routine postoperative rejection monitoring, or when rejection is suspected, blood and midstream urine samples are collected from the patients at any postoperative time.
[0026] Step 2: DNA extraction and next-generation sequencing.
[0027] Plasma cfDNA, urine cfDNA, and peripheral blood leukocyte gDNA were extracted from plasma, urine supernatant, and peripheral blood leukocyte fractions, respectively. Library construction and targeted next-generation sequencing were performed to obtain the total number of reads, allele reads, and frequency of variant alleles at SNP sites in each sample.
[0028] Step 3: Screen for informative SNP sites.
[0029] Using urinary cfDNA as the recipient enrichment reference and peripheral blood leukocyte gDNA as the allo-HSCT donor genotype reference, the frequencies of variant alleles at SNP sites were compared between the two. SNP sites with VAF less than 10% or greater than 90% in urinary cfDNA and peripheral blood leukocyte gDNA, and with presumed concordant genotypes, were screened. If these sites showed additional low-frequency allele signals in plasma cfDNA that differed from the common background pattern of both recipients and allo-HSCT donors, these sites were considered informative SNP sites, and these additional low-frequency allele signals were used as candidate signals for lung transplant donors.
[0030] Step 4: Quantification of lung-derived dd-cfDNA.
[0031] A binary mixture model was established based on the total number of informative SNP reads in plasma cfDNA and the number of candidate allele reads from lung transplant donors. Since this method does not pre-obtain lung transplant donor genotypes, the prior probability of a lung transplant donor being homozygous or heterozygous at each informative SNP locus was calculated based on publicly available allele frequencies in East Asian populations, and this prior probability was incorporated into the model. Subsequently, the proportion of lung-derived dd-cfDNA was calculated using maximum likelihood estimation, and the results are expressed as a percentage.
[0032] Step 5: Output the results.
[0033] The percentage of lung-derived dd-cfDNA output is used as an indicator for non-invasive monitoring of rejection in lung transplant patients after allo-HSCT. The test results can be comprehensively evaluated in conjunction with the patient's clinical presentation, imaging findings, pulmonary function, infection markers, and other examination results.
[0034] Example 1 Multi-source sample collection and processing are detailed below: This embodiment illustrates the methods for obtaining plasma cfDNA, urine cfDNA, and peripheral blood leukocyte gDNA in this invention.
[0035] When allo-HSCT lung transplant patients undergo routine postoperative rejection monitoring, or when clinical suspicion of potential rejection is raised, peripheral blood and midstream urine samples are collected. Peripheral blood samples are used to separate plasma and peripheral blood leukocyte components, and urine samples are used to obtain urine supernatant.
[0036] In one specific embodiment, 8 mL of peripheral blood and 8 mL of morning midstream urine were collected and placed in separate cfDNA collection tubes. The peripheral blood sample was first centrifuged at 1,600 × g for 10 minutes to separate the upper plasma layer and the lower cellular component. The separated upper plasma layer was then centrifuged at 16,000 × g for 10 minutes to remove residual cell debris, obtaining the plasma supernatant. The white membrane layer located between the plasma layer and the red blood cell layer was collected as the peripheral blood leukocyte component. The urine sample was centrifuged at 3,000 × g for 10 minutes to collect the urine supernatant and remove cellular components.
[0037] The above steps yield three types of sample sources for subsequent testing: plasma supernatant, urine supernatant, and peripheral blood leukocyte fraction. Plasma supernatant is used to extract plasma cfDNA, urine supernatant is used to extract urine cfDNA, and peripheral blood leukocyte fraction is used to extract peripheral blood leukocyte gDNA.
[0038] Example 2 DNA extraction, library construction, and targeted next-generation sequencing are detailed below: This embodiment illustrates the method for obtaining SNP site sequencing data in this invention.
[0039] DNA was extracted from the plasma supernatant, urine supernatant, and peripheral blood leukocyte fraction obtained in Example 1. Specifically, plasma cfDNA was extracted from the plasma supernatant, urine cfDNA was extracted from the urine supernatant, and peripheral blood leukocyte gDNA was extracted from the peripheral blood leukocyte fraction.
[0040] In one specific embodiment, a circulating cell-free nucleic acid extraction kit was used to extract plasma cfDNA and urine cfDNA from the obtained plasma supernatant and urine supernatant, respectively, according to the kit's instructions. A blood genomic DNA extraction kit was used to extract peripheral blood leukocyte gDNA according to the kit's instructions. After quality control (including DNA concentration detection and fragment length analysis, which was outsourced to a biotechnology company), 30 ng of cfDNA / gDNA was used for sequencing library construction. After library construction, the library concentration, fragment length distribution, and effective library ratio were quality controlled. After passing quality control, sequencing was performed on the Illumina sequencing platform using paired-end 150 bp (PE150) mode, with a target effective sequencing data volume of 10 ± 5 million reads. After quality control of the raw sequencing data, adapter sequences, low-quality reads, and duplicate reads were removed, and sequencing reads meeting the analysis requirements were retained for subsequent SNP identification and dd-cfDNA quantification analysis.
[0041] Through the above steps, targeted sequencing data were obtained from three types of samples: plasma cfDNA, urine cfDNA, and peripheral blood leukocyte gDNA. The targets were SNP sites known in current technology to be linked to heterologous cfDNA.
[0042] Example 3 Sequence alignment and SNP site information extraction are detailed below: This embodiment illustrates how the total number of reads, allele reads, and variant allele frequencies for each SNP locus are obtained in this invention.
[0043] The raw sequencing reads obtained in Example 2 underwent quality control, and the quality-controlled reads were aligned to the human reference genome. After alignment, duplicate reads were removed, and low-quality reads were filtered out. Subsequently, variant identification and deep allele extraction were performed within the target SNP regions.
[0044] In one specific embodiment, BWA-MEM was used to align the raw sequencing reads to the GRCh38 reference genome. Picard and Sambamba were used to remove duplicate reads and filter low-quality reads. GATK was used to identify variants within the target region. For each target SNP locus, the total number of reads, the number of reference allele reads, and the number of surrogate allele reads were extracted from plasma cfDNA, urine cfDNA, and peripheral blood leukocyte gDNA, and the variant allele frequency (VAF) was calculated using Equation I.
[0045] Allele frequency of variants = Alt.dp / All.dp (Equation I); Alt.dp represents the number of allele substitution reads, and All.dp represents the total number of reads.
[0046] By following the steps above, we can obtain the data foundation required for subsequent informative SNP screening.
[0047] Example 4 Screening for informative SNP sites, as detailed below: This embodiment is used to illustrate the identification method of candidate signals for lung transplant donors in this invention.
[0048] For the SNP locus information obtained in Example 3, urine cfDNA was used as the recipient enrichment reference, and peripheral blood leukocyte gDNA was used as the donor genotype reference for allo-HSCT. For each SNP locus, the frequency of variant alleles in urine cfDNA and peripheral blood leukocyte gDNA was calculated using Formula I.
[0049] When the frequency of a variant allele at a certain SNP locus in urinary cfDNA is less than 10% or greater than 90%, the recipient is considered to have a homozygous genotype at that locus (presumed genotype). When the frequency of a variant allele at a certain SNP locus in peripheral blood leukocyte gDNA is less than 10% or greater than 90%, the allo-HSCT donor is considered to have a homozygous genotype at that locus (presumed genotype).
[0050] Further screening was conducted for SNP sites where the recipient's presumed genotype matched that of the allo-HSCT donor. For these sites, the recipient-derived DNA and allo-HSCT donor-derived DNA shared the same background allele pattern. If an additional low-frequency allele signal, different from the aforementioned common background pattern, was detected in plasma cfDNA at this site, the SNP site was designated as an informative SNP site, and this additional low-frequency allele signal was considered a candidate signal for lung transplant donor origin.
[0051] Through the above screening, informative SNP sites for quantification of lung-derived dd-cfDNA can be obtained from the mixed background of plasma cfDNA without prior knowledge of lung transplant donor genotype information.
[0052] Example 5 The binary mixed model was used to calculate lung-derived dd-cfDNA, as follows: This embodiment is used to illustrate the calculation method of the proportion of lung-derived dd-cfDNA in this invention.
[0053] For the informative SNP loci screened in Example 4, the total number of reads and the number of candidate allele reads from lung transplant donors were extracted from plasma cfDNA at each locus. Let f be the proportion of lung-derived dd-cfDNA. Since this method does not pre-obtain the lung transplant donor genotype, the lung transplant donor may be homozygous or heterozygous at each informative SNP locus. Based on publicly available population allele frequencies, the prior probability of the lung transplant donor being homozygous or heterozygous at each informative SNP locus was calculated. The prior probability was obtained based on allele frequencies from publicly available population databases (e.g., the East Asian population from the 1000 Genomes Project) and calculated according to Hardy-Weinberg equilibrium. Let p be the frequency of the substitute allele at a certain locus, and q = 1 - p be the reference allele frequency; then the prior probabilities of homozygosity and heterozygosity are q and q, respectively. 2 or p 2 And 2pq.
[0054] In one specific embodiment, the aforementioned prior probabilities are calculated based on the allele frequencies of the East Asian population from the 1000 Genomes Project. Subsequently, based on the total number of reads at each informative SNP locus, the number of candidate allele reads, and the prior probabilities of homozygous or heterozygous morphology in lung transplant donors, a binomial mixture model is established to calculate the proportion of lung-derived dd-cfDNA. Then, using the proportion of lung-derived dd-cfDNA, f, as the parameter to be estimated, maximum likelihood estimation is used to determine its optimal value; in one specific embodiment, a quasi-Newton method can be used to numerically optimize the log-likelihood function to obtain an estimated value of f. Finally, the percentage of f, i.e., the level of lung-derived dd-cfDNA, is output. Its likelihood function is defined as: , in, and For each informative site Total number of reads, number of second allele reads; and These represent the prior probabilities that a lung transplant donor is homozygous or heterozygous at this locus, calculated based on allele frequencies in East Asian populations from the 1000 Genomes Project. Let be the binomial probability mass function, where p This represents the theoretical probability that a sequencing read is a minor allele; in this embodiment, the donor is homozygous. p = f When the donor is heterozygous p = f / 2. The optimal parameter f, i.e. the proportion of lung-derived dd-cfDNA, was calculated by maximizing the log-likelihood function using the quasi-Newton method.
[0055] The above steps enable the quantification of lung graft-derived cfDNA without relying on the lung transplant donor's genotype.
[0056] Example 6 Clinical sample testing and grouping verification are detailed below: This embodiment illustrates the application and detection effect of the present invention in clinical samples from patients who underwent lung transplantation after allo-HSCT.
[0057] Patients who had previously undergone allo-HSCT and subsequently received a lung transplant were included. Peripheral blood and urine samples were collected during testing, and lung-derived dd-cfDNA levels were calculated according to the methods described in Examples 1 to 5.
[0058] In one specific embodiment, 20 patients who underwent double lung transplantation after allo-HSCT were included, and a total of 49 postoperative dd-cfDNA testing timepoint samples were obtained. Based on their clinical status at the time of sampling, the samples were divided into a stable group and an acute rejection group. The stable group consisted of 26 samples, and the acute rejection group consisted of 23 samples.
[0059] The stable group was defined as lung transplant recipients who were in a stable state for more than 30 days post-transplantation and without acute rejection, infection, or other definite complications within 2 weeks before and after sampling. Acute rejection was determined based on a comprehensive assessment of clinical manifestations, imaging changes, decreased lung function, response to hormone therapy, and exclusion of infection and other complications, and was independently evaluated by a clinician. Existing detection methods (see CN201710334862.1) were used as controls.
[0060] The results are as follows Figure 2 As shown, using existing detection methods, there is significant overlap in the distribution of dd-cfDNA between the stable group and the acute rejection group, with no significant difference between the groups. After source correction using the method provided in this invention, the level of lung-derived dd-cfDNA is increased in the acute rejection group, while it is mainly distributed in the lower range in the stable group. The average proportion of lung-derived dd-cfDNA in the stable group is 0.96%, and the average proportion of lung-derived dd-cfDNA in the acute rejection group is 2.07%, with a statistically significant difference between the two groups.
[0061] The above results demonstrate that the present invention can separate lung graft-derived injury signals in the context of three-source cfDNA, thereby improving the interpretability of dd-cfDNA detection results in lung transplant patients after allo-HSCT.
[0062] Example 7 Threshold determination and detection performance evaluation are detailed below: This embodiment is used to illustrate the threshold determination and discrimination performance of the output results of the present invention.
[0063] The receiver operating characteristic (ROC) curve was used to analyze the ability of dd-cfDNA to distinguish between acute rejection and stable state before source correction (existing detection methods) and after source correction (the detection method provided by this invention), and the Youden index was used to determine the reference cutoff value.
[0064] In one specific embodiment, the result is as follows: Figure 3 As shown in Figure A, the area under the ROC curve (AUC) of the existing detection method for dd-cfDNA in differentiating between acute rejection and stable state is 0.66. Using the detection method provided by this invention, the AUC of lung-derived dd-cfDNA is improved to 0.93. The precision-recall curve is shown below. Figure 3As shown in B, the area under the uncorrected precision-recall curve (AUPRC) is 0.59, which increases to 0.95 after source correction.
[0065] When using the Youden index to determine the cutoff value, the optimal cutoff value for dd-cfDNA in existing detection methods is 3.6%. Based on the model and samples provided by this invention, the sensitivity, specificity, and accuracy calculated using statistical formulas are 56.5%, 76.9%, and 67.3%, respectively. After source correction, the optimal cutoff value for lung-derived dd-cfDNA is 1.5%, with sensitivity, specificity, and accuracy of 82.6%, 100.0%, and 91.8%, respectively.
[0066] The above results demonstrate that the present invention improves the ability of dd-cfDNA to detect acute rejection by three-source correction and can provide a reference cutoff value for lung-derived dd-cfDNA applicable to lung transplant patients after allo-HSCT.
[0067] Example 8 The detection performance was evaluated using the previous 1.17% threshold, as follows: This embodiment is used to illustrate the improvement effect of the present invention on the interpretability of dd-cfDNA threshold in previous lung transplants.
[0068] In conventional lung transplant recipients, 1.17% can be used as a reference threshold for acute rejection-related dd-cfDNA. However, when this threshold is applied to lung transplant patients after allo-HSCT, the specificity and accuracy of existing dd-cfDNA detection methods are affected by mixed background from three sources.
[0069] In one specific embodiment, when using a 1.17% threshold, the results are shown in Table 1. The sensitivity, specificity, and accuracy of the uncorrected dd-cfDNA were 91.3%, 30.8%, and 59.2%, respectively. After source correction using the method of the present invention, at the same 1.17% threshold, the sensitivity, specificity, and accuracy were 91.3%, 69.2%, and 79.6%, respectively.
[0070] Table 1. Evaluation results of detection performance using the previous 1.17% threshold.
[0071] The above results demonstrate that the system provided by this invention can not only provide new reference cutoff values, but also improve the interpretability of previous lung transplant dd-cfDNA thresholds in lung transplant patients after allo-HSCT.
[0072] Comparative Example 1 Traditional dd-cfDNA analysis without three-source correction is as follows: This comparative study adopted the traditional lung transplant dd-cfDNA analysis approach, calculating dd-cfDNA levels based solely on plasma cfDNA, without introducing urinary cfDNA as a recipient enrichment reference, and without using peripheral blood leukocyte gDNA to characterize allo-HSCT donor genotypes.
[0073] The results from the samples used in the examples showed that, in lung transplant patients after allo-HSCT, there was significant overlap between the stable and acute rejection groups for uncorrected dd-cfDNA, with no significant differences between the groups. The AUC and AUPRC for uncorrected dd-cfDNA in differentiating between acute rejection and stable status were 0.66 and 0.59, respectively. The optimal cutoff value determined using the Youden index was 3.6%, with a sensitivity of 56.5%, a specificity of 76.9%, and an accuracy of 67.3%.
[0074] The above results indicate that traditional uncorrected dd-cfDNA analysis methods are inadequate for handling the background of cfDNA from three sources in lung transplant patients after allo-HSCT, and the interpretability of the test results is limited.
[0075] Example 9 The dd-cfDNA analysis using the three-source correction method of this invention and the quantitative verification of lung-derived dd-cfDNA based on lung tissue genotype reference are as follows: This embodiment employs the method provided by the present invention to jointly analyze plasma cfDNA, urine cfDNA, and peripheral blood leukocyte gDNA. Urinary cfDNA is used as the recipient enrichment reference, and peripheral blood leukocyte gDNA is used as the donor genotype reference for allo-HSCT. Additional low-frequency allele signals are identified in plasma cfDNA, and lung-derived dd-cfDNA is calculated using a binary mixture model established for calculating the proportion of lung-derived dd-cfDNA.
[0076] This embodiment is used to verify the reliability of the lung-derived dd-cfDNA quantification results obtained by the present invention without prior knowledge of the lung transplant donor's genotype information. For samples with lung tissue specimens, DNA is extracted from the lung tissue and SNP detection is performed to obtain the genotype information of the lung transplant donor at the corresponding SNP loci. Using the lung transplant donor genotype obtained from lung tissue DNA as a reference, the proportion of lung-derived dd-cfDNA is calculated, and a correlation analysis is performed with the lung-derived dd-cfDNA quantification value obtained by the method of the present invention without a lung transplant donor genotype reference.
[0077] The results are as follows Figure 4As shown, the quantitative value of lung-derived dd-cfDNA obtained by the method of the present invention is significantly positively correlated with the quantitative value of lung-derived dd-cfDNA obtained based on lung tissue genotype reference. The Pearson correlation coefficient is 0.985; the linear fitting equation is: lung-derived dd-cfDNA (%) obtained by the method of the present invention = 0.897 × lung-derived dd-cfDNA (%) obtained by the lung tissue genotype reference method - 0.247, and the coefficient of determination R0 2 =0.971. The above results indicate that, without prior knowledge of the lung transplant donor's genotype, the lung-derived dd-cfDNA quantification results obtained by the method of the present invention are highly correlated with the quantification results obtained under conditions where the lung transplant donor's genotype is known, supporting the reliability of the method of the present invention for lung-derived dd-cfDNA quantification.
[0078] Compared with Comparative Example 1, the method provided by this invention improves the ability to identify and quantify lung-derived dd-cfDNA in the context of three-source cfDNA, and improves the ability to distinguish the rejection monitoring results of lung transplant patients after allo-HSCT.
[0079] The above embodiments and comparative examples demonstrate that, by jointly analyzing plasma cfDNA, urine cfDNA, and peripheral blood leukocyte gDNA, this invention can identify and quantify lung graft-derived dd-cfDNA in the context of three-source cfDNA in lung transplant patients after allo-HSCT. Compared with conventional analysis methods without source correction, this invention can improve the ability of lung-derived dd-cfDNA detection results to distinguish acute rejection states and improve the interpretability of previous dd-cfDNA thresholds in this specific patient population.
[0080] Therefore, this invention provides, for the first time, a method that eliminates the need for prior acquisition of genotype information from the recipient, allo-HSCT donor, and lung graft donor. It enables the identification and quantification of lung-derived dd-cfDNA in lung transplant recipients after allo-HSCT, accurately predicting whether rejection will occur after allogeneic hematopoietic stem cell transplantation. This overcomes the bottleneck in obtaining donor genotype information, solves the failure problem of traditional dual-genome methods, and gives the system clinical application value. It improves the accuracy and stability of lung-derived dd-cfDNA quantification; allows for repeated sampling and dynamic monitoring, providing clinicians with richer and more continuous information, guiding individualized adjustments to immunosuppressive therapy, and improving patient prognosis.
[0081] Finally, it should be noted that the above embodiments are only used to illustrate the technical solutions of the present invention and not to limit them. Although the present invention has been described in detail with reference to preferred embodiments, those skilled in the art should understand that modifications or equivalent substitutions can still be made to the technical solutions of the present invention, and these modifications or equivalent substitutions cannot cause the modified technical solutions to deviate from the spirit and scope of the technical solutions of the present invention.
Claims
1. A donor reference free multi-genome cfDNA provenance determination system, characterized in that: The system consists of an SNP alignment analysis module, a lung-derived dd-cfDNA ratio calculation module, and a result output module.
2. The donor independent multi-genome cfDNA provenance determination system of claim 1, wherein: The SNP alignment analysis module analyzes SNP sites in quality-controlled sequencing data of plasma cfDNA, urine cfDNA, and peripheral blood leukocyte gDNA to screen out informative SNP sites.
3. The donor independent multi-genome cfDNA provenance determination system of claim 2, wherein: When analyzing SNP sites, quality-controlled reads are aligned to the human reference genome. After alignment, duplicate reads are removed and low-quality reads are filtered out. Variation identification and allele extraction are performed in the target SNP region. The target SNP region is a known SNP site linked to heterologous cfDNA.
4. The no-donor reference multi-genome cfDNA provenance determination system of claim 3, wherein, The method for screening informative SNP loci is as follows: For each target SNP locus, the total number of reads, the number of reference allele reads, and the number of alternative allele reads are extracted from plasma cfDNA, urine cfDNA, and peripheral blood leukocyte gDNA, respectively. Then, the frequency of variant alleles in urine cfDNA and peripheral blood leukocyte gDNA is calculated using the following formula, and the genotypes of the recipient and allo-HSCT donor at this SNP locus are estimated based on the frequency of variant alleles. SNP loci with the same estimated genotypes are selected as common background candidate loci. If an additional low-frequency allele signal different from the common background pattern is detected in plasma cfDNA at this site, then the SNP site is considered an informative SNP site. The frequency of variant alleles = Alt.dp / All.dp, where Alt.dp is the number of alternative allele reads and All.dp is the total number of reads.
5. The donor independent multi-genome cfDNA provenance determination system of claim 3, wherein: When the frequency of a variant allele at a certain SNP site in urinary cfDNA is less than 10% or greater than 90%, the recipient is considered to have a homozygous genotype at that site; when the frequency of a variant allele at a certain SNP site in peripheral blood leukocyte gDNA is less than 10% or greater than 90%, the allo-HSCT donor is considered to have a homozygous genotype at that site.
6. The donor independent multi-genome cfDNA provenance determination system of claim 5, wherein: The informative SNP sites are SNP sites in which the VAF in urine cfDNA is less than 10% or greater than 90%, the VAF in peripheral blood leukocyte gDNA is less than 10% or greater than 90%, and the presumed genotypes of both are consistent, and which show additional low-frequency allele signals in plasma cfDNA that are different from the common background pattern of the recipient and allo-HSCT donor.
7. The donor independent multi-genome cfDNA provenance determination system of claim 5, wherein: The lung-derived dd-cfDNA proportion calculation module calculates the prior probability that the lung transplant donor is homozygous or heterozygous at each informative SNP locus based on the allele frequency of a specific population. Then, the prior probability is incorporated into a binomial mixture model, and the log-likelihood function is maximized using the quasi-Newton method. Based on this, the maximum likelihood estimate of the lung-derived dd-cfDNA proportion is obtained.
8. The donor independent multi-genome cfDNA provenance determination system of claim 7, wherein: The quasi-Newton method for maximizing the log-likelihood function is defined as follows: , wherein, and are the total number of reads and the number of reads for the minor allele, respectively, for each informative site . and are the prior probabilities that the lung transplant donor is homozygous, heterozygous, respectively, at the site; is the binomial probability mass function, where p represents the theoretical probability that a sequencing read is a minor allele read.
9. The donor independent multi-genome cfDNA provenance determination system of claim 8, wherein: The results output module outputs the results based on the results of the lung-derived dd-cfDNA proportion calculation module. The output results are classified as low risk of rejection, high risk of rejection, greater than or equal to 1.17% as high risk of rejection, and less than 1.17% as low risk of rejection.
10. An instrument comprising the donor-free multi-genomic cfDNA tracing system according to any one of claims 1-9.
Citation Information
Patent Citations
SNP molecular markers for detecting heterologous cfDNA, detection methods, and applications
CN107254514B
