Abstract
Objective
This study aims to explore HPV genotyping in the cervical specimen using VirCapSeq by comparing the results with the reverse blot hybridization assay (REBA).
Methods
A secondary cross-sectional data of HPV genotypes in 35 cervical specimens was obtained from VirCapSeq and REBA methods. The .FASTQ files were downloaded from the NCBI Sequence Read Archive (SRA) (accession number PRJNA766412) and HPV genotyping was bioinformatically analyzed by mapping the sequences to the PaVE database. HPV genotypes detected by REBA and NGS were compared. All specimens were stratified by histology into cervical intraepithelial neoplasia grades 1 (CIN1) and 2/3 (CIN2/3).
Results
NGS via VirCapSeq detected HPV DNA in 100% of the samples, whereas the REBA (hybridization-based) assay diagnosed HPV DNA in 85.71%. While the limitation of the conventional methods for HPV genotyping is the use of primers or probes, NGS detected a broader range. The results showed that mixed infections were detected in all samples by NGS, with HPV16 and HPV52 being the most abundant genotypes.
Conclusions
HPV genome abundance, coverage, and diversity were associated with detection discrepancies between the methods, highlighting the enhanced sensitivity and diagnostic capabilities of NGS. These findings underscore the potential of NGS technologies for comprehensive HPV genotyping, advancing cervical cancer screening, and epidemiological studies. Future research should address cost barriers and expand cohort sizes to validate these findings.
Keywords
Introduction
The viral agent that is most frequently linked to cancer cases is human papillomavirus (HPV).1,2 Notably, 80% of these cases are found in women due to invasive cervical cancer (ICC). 2 In addition, it has been reported that there will be more than 660,000 new cases and 350,000 fatalities in 2022. 3 Hence, the diagnosis of HPV infection is essential to the control of HPV-associated cancer. According to the American Cancer Society Guidelines, cytology screening (Papanicolaou test, also known as the Pap test or Pap smear) is recommended as a primary method for routine screening of cervical cancer in women ages between 21–29 years old. 4 In women aged above 30 years old and older, FDA-approved high-risk human papillomaviruses (Hr-HPVs) testing or cotesting (Hr-HPVs testing and cytology) is included as the screening test every 5 years.5,6 HPV infection can be indirectly seen by finding a shrunken nucleus in large cytoplasmic vacuoles called “koilocytes”. 7 In abnormal cytology cases, cone biopsies, endocervical scraping, and colposcopy with biopsy are also carried out in order to confirm the screening results. Changes in a biopsy, which are called cervical intraepithelial neoplasia (CIN) or dysplasia, are determined and categorized as at least 3 stages of CINs, including CIN1 (mild dysplasia), CIN2 (moderate dysplasia), and CIN3 (severe dysplasia). 8 If cancerous cells are found, they will be classified as squamous cell carcinoma (SCC), adenocarcinoma, etc.
Since HPV infection is the major factor in cervical carcinogenesis, HPV DNA detection is included to increase the sensitivity of diagnosis. To this date, the FDA has approved several HPV detection systems that utilize various molecular testing methods to identify HPV infection, particularly those associated with cervical cancer (Hr-HPVs),9,10 for example, Cervista HR HPV test (Hologic/Gen-Probe, San Diego, CA, USA), 11 Cobas 4800 HPV test (Roche, Pleasanton, CA, USA),12,13 BD Onclarity HPV (Becton Dickinson, Franklin Lakes, NJ, USA), 10 and RealTime High Risk (HR) HPV (Abbott, Chicago, IL, USA). 14 Although the efficiency of these machines for HPV detection is useful, these systems can detect at least 12 −14 types of Hr-HPVs. The Cobas 4800 system can detect 14 types of Hr-HPVs, including HPV16, 18, 31, 33, 35, 39, 45, 51, 52, 56, 58, 59, 66, and 68. However, besides HPV16 and 18, Cobas cannot identify the other HPV types and reports as Hr-HPVs positive. In addition, the low-risk (Lr)-HPVs are misdiagnosed. To detect the diverse Hr- and Lr-HPVs genotypes, Reverse Blot Hybridization Assay (REBA, Molecules and Diagnostics, Wonju, Republic of Korea) has been developed to detect at least 32 types of HPV under the reverse blot hybridization assay. 15 This assay can detect 19 Hr-HPV types: HPV 16, 26, 18, 31, 33, 34, 35, 39, 45, 51, 52, 53, 56, 58, 59, 66, 68, 69, 73, and additional 13 Lr-HPV types: HPV 6, 11, 32, 40, 42, 43, 44, 54, 70, 72, 81, 84, 87.15,16 Although many methods for HPV detection have been developed, none of them can detect all HPV types. So far, more than 150 HPV genotypes have been discovered; 17 there might be more other unknown Hr-HPV types associated with cancer development. Due to the advanced technology to date, the method for detecting the viral community (virome) has been evaluated through the next-generation sequencing (NGS) era. To enhance the virome characterization capability, a new technique for virus-sequence enrichment called “virome-capture sequencing (VirCapSeq)” has been developed. 18 This method is a kind of target enrichment by positive selection that consists of over 2 million 50- to 105-mer oligonucleotides based on known viral genome sequences. 18 As a result, it is possible that more HPV genotypes might be detected through this method, and it is not necessary to design specific primers or probes specific to each HPV type for detection. Moreover, investigating the new/unknown HPV genotypes in the cervical neoplasia samples is intriguing. Hence, our aim was to explore HPV genotyping in the cervical specimen by using VirCapSeq.
Methodology
Data recruitment
The data of HPV genotype detection by the REBA 19 and raw NGS data (.FASTQ files)20,21 of each sample were acquired from the previous studies, which were approved by the Institutional Review Board (IRB) of the Faculty of Medicine, Chulalongkorn University, a WHO-certified ethics committee (COA No. 278/2019, IRB No. 042/60, 20 March 2019; COA No. 900/2019, IRB No. 456/62, 19 August 2019; COA No. 159/2020, IRB No. 760/62, 3 February 2020; and COA No. 364/2020, IRB No. 042/60, 20 March 2020). The data obtained from 35 specimens (cervical swabs collected from patients who visited the Department of Gynecological Outpatient, King Chulalongkorn Memorial Hospital, Bangkok, Thailand, during April 2019 to December 2020) were recruited in this study according to inclusion criteria (blood-free specimens from reproductive-age women (21–50 years old) with histopathology characteristics) and exclusion criteria (active infection and use of antibiotics within a month of sample collection).
REBA HPV-Id® HPV-DNA detection and genotyping
The data of HPV genotyping by REBA was recruited from Liewchalermwong, S (2022). 19 HPV genotypes in each specimen were detected by the REBA HPV-ID® assay. This assay detects and differentiates 19 types of high-risk and 13 types of low-risk HPV using a reverse blot hybridization technique, where biotinylated PCR amplicons hybridize to specific probes immobilized on a membrane strip. After hybridization and colorimetric detection, the results appear as visible bands on the strip. The interpretation is performed by comparing the developed bands to a reference template provided by the manufacturer. β-globin was used as an internal control.
Virome capture sequencing (VirCapSeq)
The raw .FASTQ files (accession number PRJNA766412) and secondary data from Sasivimolrattana T., et al., 2022 were recruited.20,21 In those studies, to acquire the .FASTQ files, ViCapSeq was performed. Briefly, total nucleic acid of the cervical samples was extracted by the Cobas 4800 system (Roche, Switzerland). Then, cDNA was constructed by Roche Evoscript Universal cDNA Master using random hexamers (Roche, Switzerland). Then, library preparation by VerCapSeq was performed following the protocol that has been previously published.20,21 This technique is a positive selection that contains approximately 2.1 M covering the genomes of 207 known viral taxa. 22 After the library preparation step, using a MiSeq Reagent Kit v3 (Illumina, San Diego, CA, USA), the 4 nM of normalized libraries were adjusted to 18 pM and loaded onto the reagent cartridge for sequencing. The libraries were sequenced using the Illumina MiSeq platform to generate 2 × 300 bp reads.
Bioinformatics analysis
FASTQ files were downloaded from the NCBI Sequence Read Archive (SRA) (accession number PRJNA766412) in order to analyze the HPV genotypes. The quality control step was performed; the primers and adapters were removed by Trimmanomic and cutadapt, respectively. 23 Poly A tail was removed by Sickle. 24 Next, low-quality bases from the sequence end were removed, and the reads were trimmed when the median base quality fell below 20 by BBDuk version 37.10. 25 Then, the host reads (hg38) were removed by bwa 26 and samtools. 27 The genotype of HPVs was identified by blastn using PaVE as a database (https://pave.niaid.nih.gov/; access on September 30, 2022). The reads were mapped to the HPV genome using bwa 26 and Bowtie2. 28 The reads of each virus were normalized to Reads Per Kilobase per Million mapped reads (RPKM). Viral diversity (Shannon and Simpson index) was calculated by PAST statistical version 4.03.
Statistics
The statistical analysis was calculated by GraphPad Prism 8 software (Dotmatics, San Diego, CA, USA). Mann-Whitney testing (non-parametric Unpaired t-test) was used to observe the significant difference of the viral abundance (RPKM), percentage of genome coverage, and alpha diversity (Shannon and Simpson index).
Results
HPV genotyping by REBA and metagenomics next-generation sequencing
HPV genotyping and histological examination data of the 35 cervical samples were retrieved from the previous studies.19–21 The average age of the studied group was 37.971 ± 1.459 (Mean ± SEM), ranging from 23–50 years. The samples were divided into two groups based on the histology, including CIN1 (N = 21) and CIN2/3 (N = 14). Of those samples, 85.71% (N = 30) and 100% (N = 35) of the cases were found to be HPV DNA positive by REBA and NGS, respectively (Table 1). The HPV genotype in each sample detected by REBA and NGS was shown in Supplementary Table 1. HPV16 was the most ubiquitous among samples when detected by both methods (Figure 1A and B). Twenty-two samples (62.86%) were diagnosed as HPV16-positive by REBA, while 26 (74.29%) HPV16-positive samples were found by NGS (Table 1). Of the samples that tested positive for HPV16 using these two methods, 17 out of 30 samples (56.67%) demonstrated consistency in HPV16 detection. Interestingly, the same detection rate (78.57%) of HPV16 detection was found in the CIN2/3 group when diagnosed by both methods (Table 1). Unexpectedly, 5 HPV-negative by REBA were all HPV16-positive detections by NGS. We rechecked the results from the routine laboratory database using Cobas 4800 system (an FDA-approved real-time PCR-based HPV detection system), and the results were all negative, similar to REBA results. However, we hypothesized that it might be a sensitivity issue. Then, the detection of HPV by NGS was validated by comparing the results of the same samples analyzed using the Virus Identification Pipeline (VIP), a de novo assembly-based method that has been previously published.21,29 The results indicated that all of those HPV16-negative by PCR and REBA were HPV16-positive when analyzed by VIP. In addition, the presence of HPV genome was confirmed again by exporting the suspected HPV16 reads and mapping them with the reference sequence of HPV16 (Accession: K02718) using CLC Genomic Workbench 22. This suggested that the HPV detection by NGS was reliable. Besides HPV16, the REBA detected HPV52 as the second most prevalent genotype, followed by HPV42, HPV56, and HPV11 (Figure 1A). In contrast, by NGS, HPV56-positive samples (26 samples) were identified at the same rate as HPV16, followed by HPV58, HPV51, HPV66, and HPV52 (Figure 1B). In the heat map analysis, HPV16 was the most abundant among 35 samples, followed by HPV52, HPV51, HPV43, HPV58, HPV56, and others (Figure 1C). Remarkably, mixed infection of HPV genotypes was found in all samples by NGS detection (Figure 1C). Lr-HPVs co-infection was found in 9 cases (25.71%) by REBA, whereas almost all cases (34 samples; 97.14%) were identified by NGS (Table 1). HPV genotypes detected by both REBA and NGS-Metagenomics were types 6, 11, 16, 33, 39, 42, 43, 45, 51, 52, 53, 54, 56, 58, 66, 68, 73, and 87 (Figure 2). Interestingly, discrepancy genotypes between these 2 methods were demonstrated. HPV59, 72, 81, and 84 were only found by REBA, whereas HPV30, 32, 34, 35, 44, 61, 62, 67, 70, 71, 74, 82, 83, 86, 89, 90, 91, 106, and 114 were detected by NGS.
HPV genotype detection by REBA and NGS. Number of each HPV genotype positive sample by REBA (A) and NGS (B). Heat maps (C) showed the abundance of each HPV genotype by NGS. Reads Per Kilobase per Million mapped reads (RPKM) represented the viral abundance. Red color represents Hr-HPV, and blue represents Lr-HPV.
Venn diagram showed unique and shared HPV genotypes detected by REBA and NGS. (Red = Hr-HPVs; Blue = Lr-HPVs).
HPV genotyping by REBA and NGS (VirCapSeq), stratified histologically in cervical samples.
Of these 35 samples, 21 samples (60%) were CIN1, whereas 14 samples (40%) were CIN2/3 (Table 1). No difference in ages between CIN1 (38.810 ± 1.850 years) and CIN2/3 (36.714 ± 2.410 years) was found (p = 0.1577; Mann-Whitney U test). Interestingly, 5 samples (23.81%) in the CIN1 group were identified as HPV-negative by two diagnostic tests, i.e. REBA and Cobas 4800, but positive by NGS. Table 1 showed the detection rate of the top six HPV types among 35 samples (by abundance); the detection rate of each type by REBA was lower than that of NGS. The detection rates of Hr-HPVs by REBA and NGS were 76.19% (16/21) and 100% (21/21) in CIN1, and 100% in CIN2/3 by both methods. The single infection of HPV16 was only found by the REBA detection (25.71%); nevertheless, no single infection was found when performed by NGS. For the Lr-HPVs detection, the detection rates by REBA were 23.81% (5/21) and 28.57% (4/14) in CIN1 and CIN2/3, while they were 100% (21/21) and 92.86% (13/14), respectively, by NGS.
Association of genomic properties and HPV detection
To determine the impact of genomic properties on HPV detection, HPV abundance (Reads Per Kilobase per Million mapped reads; RPKM) and genome coverage were analyzed (Figure 3). We further focus on the detection of HPV16 and 52 since these types had the highest abundance of HPV reads among HPV types (Figure 1C). The samples were divided into two groups based on the REBA and NGS results: the double-positive group (REBA+, NGS+) and the NGS-positive group (REBA‒, NGS+). The results showed that the abundance of HPV16 and HPV52 in the double-positive group was higher than that in the NGS-positive alone group, especially in HPV52 (p = 0.0072) (Figure 3A). As well as in the percentage of genome coverage, the value in the double-positive group was higher than that of the NGS-positive group, especially in HPV52 (p = 0.0018) (Figure 3B). When the samples were divided according to their histological characteristics, similar results were observed (Figure 3C and D). A significant difference in HPV52 abundance and coverage was observed between double-positive and NGS-positive in the CIN1 group (p = 0.0121 and p = 0.0121, respectively), whereas a trend of difference was found in the CIN2/3 group (Figure 3C and D).
HPV abundance (RPKM) and genome coverage (%) in each group. (A) HPV abundance and (B) percentage of HPV genome coverage between the double-positive group and the NGS-positive group. (C) HPV abundance and (D) percentage of HPV genome coverage between the double-positive group and the NGS-positive group, stratified histologically. Error bars indicate the standard error of mean (SEM).
In addition to the abundance and genome coverage, the effect of HPV diversity on HPV genotyping was also evaluated. The results revealed that HPV diversity in double-positive samples was lower than that in NGS-positive samples in all circumstances (Figure 4). The statistical difference in alpha diversity of HPVs (Shannon and Simpson index) was found between HPV16 positive groups (Figure 4A and B). Similar trends were also found when the samples were divided by histological characteristics (Figure 4C and D).
HPV diversity in each group. (A-B) HPV diversity (Shannon and Simpson index, respectively) in the double-positive group and NGS-positive group. (C-D) HPV diversity (Shannon and Simpson index, respectively) in the double-positive group and the NGS-positive group, stratified histologically. Error bars indicate the minimum and maximum values.
Discussion
Detection of HPV DNA in cervical specimens is a critical tool in cervical cancer screening and prevention. Certain Hr-HPV types are strongly associated with the development of cervical intraepithelial neoplasia (CIN) and cervical cancer. HPV DNA testing helps detect those oncogenic types early. However, there is no routine laboratory diagnostic platform that can detect all types of HPV, especially Hr-HPVs. As a result, HPV DNA detection and genotyping by VirCapSeq based on the NGS platform were performed. The percentage of HPV-positive samples by NGS was 100%, whereas it dropped to 85.71% via REBA detection. Only five samples (14.28%) were reported as HPV-negative by REBA (Table 1) and confirmed HPV-negative by Cobas 4800, another diagnostic test. The percent consistency of HPV16 detection between REBA and NGS was quite low (56.67%). This might be caused by HPV16-positive detection by NGS in HPV-negative REBA samples. Similar results were also demonstrated in previous studies.21,30 Due to the high technology of NGS, the sensitivity of this technique is higher than that of routine laboratory diagnostic procedures. Furthermore, compared to NGS, the REBA method might have relatively low sensitivity because it is a semiquantitative approach that is interpreted by visually observing the band. In this study, HPV16 was the most frequent genotype detected by both REBA and NGS, consistent with its established role as the primary oncogenic type in cervical cancer. However, NGS detected HPV16 in a higher percentage of samples (74.29%) compared to REBA (62.86%) (Table 1). Similarly, other high-risk types such as HPV56, HPV58, and HPV51 were more frequently detected by NGS, highlighting its enhanced capacity to identify less abundant genotypes. A single infection was not observed by NGS; on the other hand, co-infections with multiple HPV genotypes were found in all samples (Table 1 and Supplementary Table 1). As predicted, most of them, such as HPV16, 51, 52, 56, and 58, are the genotypes that have been circulated among Thai women, 31 suggesting that it is possible to detect some of these genotypes by NGS, but the misdiagnosis by REBA might be due to the sensitivity issue. Although our study showed that 100% of multiple HPV infections (by NGS) were detected, this prevalence does not represent the typical population in this region because the sample used in these studies was initially chosen based on certain criteria. This study revealed that several HPV genotypes (HPV30, 34, 35, 67, and 82) were detected only by NGS (Figure 2). This implies that there are other Hr-HPVs, such as HPV67 and 82, in the samples that were misdiagnosed by REBA. Unexpectedly, HPV34 and 35 were unable to be detected by REBA, but NGS can detect both (Figure 1A). Especially HPV35, which ranked as one of the top six HPVs that is usually found in cervical cancer samples in Thailand. 31 Due to the high sensitivity of NGS, very low relative abundance of those HPV types could be detected (Figure 1C). By NGS, mixed HPV infections were identified in all samples, whereas REBA detected HPV16 single infection in 25.71% of cases (Table 1). This disparity highlights the benefits of NGS in detecting several HPV genotypes, even in low-abundance infection situations. This enhances diagnostic precision and offers a more thorough comprehension of HPV diversity in cervical lesions. The HPVs with the highest abundances found in our investigation were HPV16 and 52, as seen in Figure 1C. This finding is in line with a prior study that found HPV16 and 52 were commonly found in low- or high-grade lesions in Thai people. 31 Consequently, this validated the accuracy of the HPV genotyping NGS results.
HPV16 infection might be the tip of the iceberg of cervical carcinogenesis since several other HPV genotypes were found in the HPV16-positive samples. Hence, these findings support the integration of NGS into cervical cancer screening and research, with the potential to enhance our understanding of HPV epidemiology and its role in cervical lesion progression since this technique can detect more genotypes at the same time. In addition, it is possible that the new candidate of Hr-HPV will also be discovered as a result of HPV detection by NGS. However, although NGS has a high sensitivity for detection, several limitations of the uses of NGS for diagnosis, such as costs, procedures, and post-analysis interpretation, are found. Hence, NGS for HPV screening might be used for diagnosis for some specific objectives, for example, to detect the presence of new/unknown/uncommon HPV genotypes, and may be in samples of cancers that showed negative HPV detection using the conventional assays.
A previous study showed that REBA is a sensitive test for genotyping HPV in clinical specimens, 15 and the cost per sample and turnaround are lower than those of NGS. Our results showed that routine laboratory diagnosis for HPV detection not only misdiagnosed (HPV negative detection) but also did not cover all the HPV genotypes associated with cervical cancer development. NGS might be additional confirmed tests to detect some HPV-negative samples or confirm the presence of specific genotypes. Several methods for HPV detection based on the NGS platform have now been developed, for example, HPV-meta, 32 and ChapterDx HPV-STI NGS assay. 30 Since NGS can detect a broad range of Hr-HPVs and Lr-HPVs, the benefit to patients from cervical cancer screening will be enhanced. In addition, not only for HPV detection, this method can be used for the detection of other viral infections at the genital tracts, such as herpes simplex virus (HSV) and molluscum contagiosum virus (MCV).
As mentioned above, the contradictory results between NGS and REBA might be associated with the sensitivity issue. Previous studies demonstrated that when comparing the REBA method for the detection of HPV in known-HPV-positive samples to other different commercial kits for HPV genotyping, including the real-time PCR-based method (Anyplex II HPV28 Detection) and Chip-based technique (HPVDNAChip), the number of HPV-positive samples by REBA was lower than those methods. 33 Moreover, in another study, HPV genomes were detected by the ChapterDx HPV-STI NGS assay, while the real-time PCR-based Cobas 4800 showed negative results for HPV detection. 30 This suggested that NGS has superior sensitivity for HPV detection. Interestingly, all of the contradicted samples for HPV detection in that study were in negative for Intraepithelial lesion or malignancy (NILM) and low-grade squamous intraepithelial lesion (LSIL) groups. 30 Similar to our findings, most REBA-NGS + for HPV16 were usually found in the lower stage of cervical neoplasia (N = 6 in CIN1, N = 2 in CIN2/3 groups) (Figure 3C).
The potential factors that might affect the detection of HPV genotypes, for example, HPV abundance (RPKM) and percentage of HPV genome coverage. As predicted, a positive correlation between high HPV abundance and HPV-positive by REBA detection was found (Figure 3A and C). The HPV16 and HPV52 abundances in the REBA − NGS + group were lower when compared to those in the REBA + NGS + group (Figure 3A and C). A similar trend between high HPV genome coverage and REBA detection was found (Figure 3B and D). Hence, these results suggested that HPV abundance and genome coverage may influence the sensitivity of REBA. This trend was consistent across both histological groups, emphasizing the importance of viral load and genome integrity in influencing detection efficiency. Besides the HPV abundance and coverage, the impact of HPV diversity for HPV genotyping was also observed since mixed infection was found in all samples when diagnosed by NGS. The results revealed that high HPV diversity was usually found in the REBA negative group (Figure 4). This indicates that lower HPV diversity may enhance the detectability of specific genotypes by REBA, whereas the broader capability of NGS to detect multiple genotypes highlights its potential as a more comprehensive diagnostic tool.
Conclusions
This study demonstrates the capability of NGS-metagenomics in detecting and characterizing HPV infections, particularly in cases of mixed infections and low-abundance genotypes. This is the first study suggesting that HPV genotyping might be affected by HPV abundance, genome coverage, and diversity. Nevertheless, future study is needed to confirm the association of these factors for each kind of HPV diagnosis. Moreover, this study was limited by the small sample size. Future studies should include larger cohorts to validate these findings and explore the clinical significance of co-infections and less common genotypes. There is a need to enhance the number of samples and include cancer samples in further research. Moreover, to observe the potential of the use of NGS in other microbial detections, a comparison of the detection of other pathogens, e.g. HIV, HSV, EBV, MCV, Neisseria gonorrhoeae, and Chlamydia trachomatis, in cervical specimens between NGS and other methods is warranted.
Supplemental Material
sj-docx-1-sci-10.1177_00368504251334515 - Supplemental material for Virome capture sequencing for comprehensive HPV genotyping in cervical samples
Supplemental material, sj-docx-1-sci-10.1177_00368504251334515 for Virome capture sequencing for comprehensive HPV genotyping in cervical samples by Thanayod Sasivimolrattana, Sasiprapa Liewchalermwong, Wasun Chantratita, Insee Sensorn, Arkom Chaiwongkot and Parvapan Bhattarakosol in Science Progress
Footnotes
Acknowledgment
We would like to thank Miss Apaporn Rodpan, Nextgen Network Corporation Co., Ltd for her kind help in bioinformatics analysis.
ORCID iDs
Ethical considerations
This study was conducted using secondary data from the projects approved by the Institutional Review Board of the Faculty of Medicine, Chulalongkorn University, a WHO certified ethics committee (COA No. 278/2019, IRB No. 042/60, 20 March 2019; COA No. 900/2019, IRB No. 456/62, 19 August 2019; COA No. 159/2020, IRB No. 760/62, 3 February 2020; and COA No.364/2020, IRB No. 042/60, 20 March 2020). No personally identifiable information was included. Thus, this study did not have to apply for ethics approval nor informed consent.
Author contributions/CRediT
Conceptualization, P.B. and T.S.; methodology, T.S., S.L., A.C., P.B.; software, P.B. and T.S.; validation, P.B. and A.C.; formal analysis, T.S. and S.L.; investigation, T.S., I.S., W.C., A.C., P.B.; resources, P.B., A.C. W.C.; data curation, T.S.; writing—original draft preparation, T.S.; writing—review and editing, P.B., A.C., I.S.; visualization, T.S.; supervision, P.B.; project administration, P.B.; funding acquisition, P.B. and A.C. All authors have read and agreed to the published version of the manuscript.
Funding
The authors disclosed receipt of the following financial support for the research, authorship, and/or publication of this article: This research was funded by a grant from the Government Health Systems Research Institute (HSRI; grant number 61-074 and 63-084), the Ratchadaphiseksomphot Matching Fund (grant number RA-MF-11/63) and the Thailand Science Research and Innovation Fund (TSRI) (grant number CU_FRB640001_01_30_9).
Conflicting interests
The authors declared no potential conflicts of interest with respect to the research, authorship, and/or publication of this article.
Data availability
The data that support the findings of this study are available from the corresponding author upon reasonable request. Raw sequence files from virome capture sequencing were downloaded from the NCBI Sequence Read Archive (SRA) under accession numbers PRJNA766412.
Supplemental material
Supplemental material for this article is available online.
References
Supplementary Material
Please find the following supplemental material available below.
For Open Access articles published under a Creative Commons License, all supplemental material carries the same license as the article it is associated with.
For non-Open Access articles published, all supplemental material carries a non-exclusive license, and permission requests for re-use of supplemental material or any part of supplemental material shall be sent directly to the copyright owner as specified in the copyright notice associated with the article.
