Broad germline variant spectrum revealed by whole-exome sequencing in an underrepresented Latin American population with hereditary breast cancer

Abstract

Background

Hereditary breast cancer (BC) accounts for a significant proportion of BC cases, yet germline variant interpretation remains limited in underrepresented populations such as those from Latin America. Most available genomic reference data derive from European and North American populations, constraining variant interpretations and clinical decision-making in other regions.

Methods

We performed whole-exome sequencing in 140 Chilean patients with clinically suspected hereditary BC. Germline variants were analyzed across canonical BC predisposition genes recommended by current NCCN guidelines and an extended set of additional cancer susceptibility genes. Variants were annotated and manually classified following ACMG/AMP guidelines, incorporating ClinGen gene-disease validity resources and cancer-specific CanVIG consensus recommendations.

Results

Expanded gene analysis resulted in an increased detection of pathogenic and likely pathogenic variants (~ 18% to ~ 24%) and reduced the proportion of patients with negative genetic results compared with canonical gene analysis alone. Pathogenic variants were most frequently identified in BRCA1 and BRCA2, followed by additional cancer susceptibility genes, with truncating variants representing the predominant alteration type. Exploratory analyses revealed a broad distribution of pathogenic variants across BRCA1 and BRCA2 gene regions.

Conclusions

This study provides a comprehensive overview of the germline variant spectrum in Chilean patients with suspected hereditary BC and demonstrates the added value of extended gene analysis beyond canonical predisposition genes. These findings contribute to population-specific genomic data and support efforts to improve variant interpretation and equitable implementation of genomic medicine in underrepresented populations.

Supplementary Information

The online version contains supplementary material available at 10.1186/s40246-026-00983-2.

Introduction

Breast cancer (BC) is the most frequently diagnosed malignancy and the leading cause of cancer-related mortality among women worldwide [1]. While most cases are sporadic, approximately 5–10% of BC cases are attributed to hereditary factors, primarily caused by germline pathogenic variants in genes involved in DNA repair, genomic stability, and cell cycle regulation [2, 3]. The most well-established predisposition genes include BRCA1 and BRCA2, which confer a high lifetime risk of breast and ovarian cancer, whereas other moderate- and low-penetrance genes such as PALB2, CHEK2, ATM, and BARD1 have also been associated with increased susceptibility [4–6]. The clinical implementation of multigene panel testing through next-generation sequencing (NGS) has substantially broadened the identification of pathogenic and likely pathogenic variants, improving diagnostic yield and risk assessment strategies in hereditary BC⁷. However, the global interpretation of germline variants remains uneven. Most variant frequency data, reference allele frequencies, and pathogenicity classifications derive predominantly from populations of European or North American ancestry [7]. This imbalance has resulted in a limited understanding of genetic variability in underrepresented populations, including those from Latin America. For instance, less than 1% of samples included in large-scale repositories such as gnomAD originate from Latin American individuals [8], leading to a disproportionately high proportion of variants of uncertain significance (VUS) in these populations [9]. This underrepresentation directly impacts the accuracy of variant interpretation, as population-specific alleles may be erroneously classified or remain unrecognized due to the lack of representative reference data.

Recent studies employing whole-exome sequencing (WES) in diverse and previously underrepresented populations have demonstrated the value of comprehensive genomic approaches for identifying both known and novel germline variants associated with cancer predisposition [10–12]. These exome-based investigations have not only expanded variant catalogs but also revealed ancestry-specific mutational spectra and potential founder effects. However, comparable large-scale analyses are still lacking in Latin America, where genetic studies of hereditary BC have largely relied on targeted BRCA1/2 testing or restricted gene panels [13, 14].

In Chile, the availability of genomic data related to hereditary BC is scarce and fragmented. Most studies have focused on BRCA1 and BRCA2, often in small cohorts [15–17], while information regarding other clinically relevant cancer predisposition genes remains limited. Moreover, the Chilean population presents a unique genetic admixture resulting from European, Indigenous, and African ancestries, potentially influencing the prevalence and spectrum of germline variants [18]. This genetic heterogeneity highlights the need for local data to support accurate variant classification and risk estimation in Chilean patients. Additionally, the lack of integration of national genomic databases or coordinated variant repositories further complicates clinical interpretation and underscores the necessity of building population-specific reference datasets. Previous studied in Latin American populations have highlighted the unique genetic architecture and admixture patterns that influence germline variant distribution, underscoring the need for population-specific genomic data to improve variant interpretation and clinical decision-making [13, 17, 18].

In this context, characterizing the germline variant landscape in individuals meeting clinical criteria for hereditary BC suspicion is crucial for elucidating population-specific mutational patterns and refining variant interpretation frameworks. Here, we present a comprehensive genomic analysis of germline variants across clinically relevant cancer predisposition genes in Chilean patients with suspected hereditary BC, aiming to assess the added value of extended gene analysis and to contribute population-specific data to improve genomic medicine implementation in Chile and Latin America.

Materials and methods

Study design and participants

A total of 140 female patients diagnosed with BC and with a familial history suggestive of hereditary cancer syndromes, including hereditary breast and ovarian cancer (HBOC), hereditary diffuse gastric cancer, Lynch syndrome, familial adenomatous polyposis, or MUTYH-associated polyposis, were selected according to the National Comprehensive Cancer Network (NCCN) Clinical Practice Guidelines in Oncology [19]. Age at diagnosis ranged from 28 to 68 years.

Participants were recruited from public and private healthcare institutions in the Metropolitan Region of Santiago, Chile. Inclusion criteria were defined according to the clinical criteria for hereditary cancer syndromes established by the NCCN guidelines. Individuals who did not fulfill these clinical criteria were excluded from the study.

DNA extraction, quantification, and quality assessment

Genomic DNA was extracted from peripheral blood samples using Exgene™ Blood SV Kit (GeneAll Biotechnology, Seoul, South Korea), following the manufacturer’s protocol. Briefly, blood samples were lysed by incubation with Proteinase K and lysis buffer at 56 °C, followed by ethanol precipitation and purification using silica membrane spin columns. After sequential washing steps, DNA was eluted in nuclease-free water. DNA concentration was measured using a fluorescence-based method with the Qubit™ 1x dsDNA High Sensitivity Assay Kit (Thermo Fisher Scientific, Waltham, MA, USA), adapted for use with the Fluo-200 fluorometer (Allsheng Instruments, Hangzhou, China) according to the validated assay protocol for this platform. DNA integrity was assessed by agarose gel electrophoresis, and samples showing intact high-molecular-weight DNA were considered suitable for downstream genomic analyses.

Whole-exome sequencing (WES)

WES was performed using a targeted capture approach. Library preparation and exome enrichment were carried out using the Whole Exome Sequencing Panel (Celemics), designed for multiplexed target enrichment of exonic regions. Genomic DNA libraries were prepared using the MGI Enzymatic Library Prep Kit with single indexing, following the manufacturer’s instructions. Exome enrichment was performed using the Whole Exome Panel Target Enrichment Kit (MGI), employing streptavidin-based capture chemistry.

Library preparation and enrichment workflows included the use of CLM polymerase, CeleMag™ Streptavidin Beads, CeleMag™ Clean-up Beads, and CeleNM™ Beads (Celemics), as well as circularization using MGIEasy Circularization Module V2.0 (MGI), enabling the generation of DNA nanoballs (DNBs) for sequencing.

Sequencing was performed on the DNBSEQ-G400 platform (MGI) using the DNBSEQ G400S High-throughput Sequencing Set, generating paired-end reads of 150 base pairs. Adapters used during library preparation corresponded to the standard MGI adapters; Adapter Forward: AAGTCGGAGGCCAAGCGGTCTTAGGAAGACAA, and Adapter Reverse: AAGTCGGATCGTAGCCATGTCGTTCTGTGAGCCAAGGAGTTG.

WES was selected as a cost-effective and clinically relevant approach to capture coding variation in cancer susceptibility genes, which currently represent the primary targets of germline testing in clinical practice.

Bioinformatic processing

Bioinformatic analysis of raw WES data was performed using the nf-core/sarek workflow. Genomic coordinates of the target regions provided by Celemics (BED format) were lifted over from the GRCh37 to the GRCh38 assembly using the Picard LiftOver tool. The corresponding chain file (hg19ToHg38.over.chain.gz) was downloaded from the UCSC Genome Browser repository (http://hgdownload.soe.ucsc.edu/). Lifted-over target regions were subsequently sorted and merged according to the GRCh38 reference genome using BEDTools [20].

Raw sequence data in FASTQ format were filtered to remove low-quality reads using FASTP [21]. Filtered reads were aligned to the GRCh38 human reference genome (AWS iGenomes) using BWA-MEM [22] with default parameters. Base quality score recalibration (BQSR) was performed using known-sites from dbSNP build 151 and the Mills and 1000 Genomes Project gold-standard indel set, as provided in the GATK resource bundle. Germline variants were called using HaplotypeCaller, and joint genotyping was performed using GenotypeGVCFs. Identified single nucleotide variants (SNPs) and insertions/deletions (indels) were recalibrated using variant quality score recalibration (VQSR) with established truth sets, including HapMap, Omni, Mills, and Phase 1.

Variant annotation

Variants were annotated for genomic position, predicted functional effect, and clinical classification using ANNOVAR (http://www.openbioinformatics.org/annovar/). When available, the MANE Select transcript was used for variant annotation [23]. Variant analysis was performed using a custom virtual panel comprising 20 genes recommended by NCCN guidelines for hereditary BC (Supplementary Table 1), as well as a secondary virtual panel including 111 other cancer-related predisposition genes (Supplementary Table 1). Gene-disease validity was assessed using ClinGen classifications where available, allowing differentiation between genes with strong, moderate, or limited evidence for breast cancer susceptibility [24].

Post-annotation, variant filtering was performed using R software (https://www.r-project.org/). Synonymous variants and variants with a sequencing depth < 15 reads per sample were removed. In addition, variants with a minor allele frequency (MAF) ≥ 1% in gnomAD exomes were also excluded. Loss-of-function variants, including frameshift and nonsense mutations, were retained. Missense variants were selected only if predicted to be deleterious by functional impact predictors, defined as either a REVEL score ≥ 0.75 or concordant predictions from at least three individual tools: SIFT, PolyPhen-2, and MutationTaster.

Variants passing allele frequency and functional impact filters were further evaluated based on ClinVar clinical classification. Pathogenic and likely pathogenic variants were retained, whereas benign, likely benign, and conflicting variants were excluded. VUS were retained if they met frequency and functional impact criteria. As a final filtering control step, all selected variants were manually inspected in the corresponding CRAM files using the Integrative Genomics Viewer (IGV) [25] to assess potential sequencing or alignment artifacts. An overview of the variant filtering workflow and the number of variants retained at each analysis step is shown in Supplementary Fig. 1.

Variant classification

Identified variants were annotated and classified using a manual curation approach following the American College of Medical Genetics and Genomics and the Association for Molecular Pathology (ACMG/AMP) 2015 guidelines, incorporating subsequent updates relevant to cancer susceptibility genes [26]. Variant classification was performed by integrating multiple lines of evidence, including variant type, predicted functional impact, population frequency, and previously reported clinical associations.

Gene-disease validity was evaluated using resources from the Clinical Genome Resource (ClinGen), with particular emphasis on curated gene-disease relationships relevant to BC susceptibility. Information from ClinGen Expert Panels was considered when available to support gene-level validity and contextual interpretation.

For genes associated with hereditary cancer predisposition, variant interpretation was further informed by the CanVIG-UK Consensus Specification for Cancer Susceptibility Genes, developed under the Association for Clinical Genomic Science (ACGS) Best Practice Guidelines for Variant Classification [27]. These cancer-specific adaptations of the ACMG/AMP framework were applied to ensure appropriate weighting of evidence in the context of cancer susceptibility genes.

Clinical variant databases, including ClinVar and LOVD, were used as supportive evidence to inform classification. However, no variant was classified solely based on database assertions. Final variant classification was determined through manual expert review, integrating all available evidence and applying ACMG/AMP criteria in a consensus-based manner.

Where available, ClinGen gene-specific variant interpretation recommendations (VCEP) were considered to further support classification. In addition, variant classifications were cross-referenced with ClinVar, CanVIG, and LOVD databases. While high concordance was observed for most pathogenic and likely pathogenic variants, final classifications were determined according to ACMG/AMP criteria.

Results

Gene selection and analytical framework

To contextualize the interpretation of germline findings, variant analyses were performed using a two-tiered analytical framework. First, variants were evaluated across the 20 breast cancer predisposition genes recommended by current NCCN guidelines, representing the standard clinical approach. Subsequently, the analysis was expanded to include an additional set of 111 cancer susceptibility genes to explore the broader landscape of germline predisposition, assess genetic pleiotropy, and evaluate the added diagnostic value of extended gene analysis in this cohort.

Overall variant landscape

A comprehensive germline variant analysis was performed in the study group, revealing a heterogeneous landscape of genetic alterations across multiple cancer predispositions genes. The overall distribution of variant types and the number of variants identified per gene are summarized in Supplementary Fig. 2. Truncating variants, including frameshift insertions, frameshift deletions, and nonsense variants, accounted for most pathogenic and likely pathogenic findings, whereas missense variants represented a smaller proportion of clinically relevant alterations. This distribution is consistent with the established molecular mechanisms underlying hereditary cancer susceptibility in high-penetrance genes.

Impact of extended gene analysis beyond canonical breast cancer genes

When the analysis was restricted to the 20 canonical BC predisposition genes recommended by NCCN guidelines, pathogenic and likely pathogenic variants were identified in a subset of patients (Fig. 1). Expanding the analysis to include an additional 111 genes associated with hereditary cancer predisposition resulted in a substantial increase in the detection of pathogenic and likely pathogenic variants (18.7% v/s 23.6%). As expected, this broader approach was also associated with an increased identification of VUS (4.3% v/s 19.3%). Importantly, the extended gene analysis markedly reduced the proportion of patients classified as having negative genetic results (77% v/s 57.1%). Notably, several pathogenic variants were identified exclusively through the extended gene analysis, highlighting the added value of a broader germline evaluation in this group of patients.

Fig. 1.

Impact of extended gene analysis on variant detection. Comparison of genetic testing results obtained using the 20 NCCN-recommended BC predisposition genes A, versus and extended set of cancer susceptibility genes, plus 20 NCCN-recommended genes B. Bars represent the proportion of patients with pathogenic (P) and likely pathogenic (LP) variants, variants of uncertain significance (VUS), and negative results

Distribution of pathogenic and likely pathogenic variants across genes

The distribution of pathogenic and likely pathogenic variants identified across the group of patients is summarized in Fig. 2. BRCA1 and BRCA2 were the most frequently altered genes, together accounting for most pathogenic findings in this cohort. Additional recurrent pathogenic variants were observed in other cancer predisposition genes, including MUTYH, ATM, ATRX, CHEK2, FANCA, FANCM, RAD51D, RECQL4, and TP53. The TP53 pathogenic variant identified in this cohort was not associated with clinical features suggestive of a classical Li-Fraumeni syndrome phenotype based on available clinical information. Most pathogenic and likely pathogenic variants correspond to truncating alterations, reinforcing their established role in cancer susceptibility. Table 1 details the specific pathogenic and likely pathogenic variants identified, including genomic annotation, variant type, ACMG classification, and the number of carriers for each variant. Oncoplot and variants detail for VUS are shown in supplementary Fig. 3 and Supplementary Table 2. Notably, variants identified in genes with limited or emerging evidence for breast cancer susceptibility were interpreted cautiously. In particular, monoallelic MUTYH variants were not considered contributory to breast cancer risk, given the recessive inheritance pattern associated with this gene.

Fig. 2.

Distribution of pathogenic and likely pathogenic germline variants across cancer predisposition genes. Oncoplot summarizing pathogenic and likely pathogenic variants identifies in this study. Genes are ordered by decreasing frequency of variants. Columns represent individual patients, and colors indicate variant types. Upper bars denote the number of variants per patient, and side bars indicate the number of variants per gene

Table 1.

Pathogenic and likely pathogenic germline variants identified in the study group

Gene	Transcript (MANE)	cDNA change (HGVS)	Protein change (HGVS)	Variant type	ACMG Classification	RS ID	Number of carriers
BRCA1	NM_007297.4	c.3331_3334del	p.Gln1111fs	Frameshift Del	P	rs80357701	4
		c.3759dup	p.Lys1254Ter	Nonsense	P	rs80357687	2
		c.3817 C > T	p.Gln1273Ter	Nonsense	P	rs80357208	2
		c.5266dup	p.Gln1756fs	Frameshift Ins	P	rs80357906	2
		c.5102_5103del	p.Leu1701fs	Frameshift Del	P	rs80357608	1
		c.5164T > A	p.Ser1722Thr	Missense	LP	rs483353100	1
BRCA2	NM_000059.4	c.5144_5147del	p.Tyr1716fs	Frameshift Del	P	rs276174854	5
		c.8987T > A	p.Leu2996Ter	Nonsense	P	rs864622200	3
		c.6275_6276del	p.Leu2092fs	Frameshift Del	P	rs11571658	1
		c.8065_8066del	p.Val2690fs	Frameshift Del	P	rs80359694	1
		c.4737_4738ins	p.Glu1581fs	Frameshift Ins	P	rs864622401	1
MUTYH	NM_001048174.2	c.1103G > T	p.Gly368Asp	Missense	P	rs36053993	3
ATM	NM_000051.4	c.7886_7890del	p.Ile2629fs	Frameshift Del	P	rs1450394308	1
ATRX	NM_000489.6	c.1330 C > T	p.Arg444Ter	Nonsense	LP	rs2148625435	1
CHEK2	NM_007194.4	c.1100del	p.Thr367fs	Frameshift Del	P	rs555607708	1
FANCA	NM_000135.4	c.2107 C > T	p.Gln703Ter	Nonsense	P	rs1555548512	1
FANCM	NM_020937.4	c.5791 C > T	p.Arg1931Ter	Nonsense	LP	rs144567652	1
RAD51D	NM_002878.4	c.216 C > A	p.Tyr72Ter	Nonsense	P	rs148690585	1
RECQL4	NM_004260.4	c.2590 C > T	p.Gln864Ter	Nonsense	P	rs1054186954	1
TP53	NM_000546.6	c.638G > A	p.Arg213Gln	Missense	P	rs587778720	1

P: Pathogenic; LP: Likely Pathogenic

Exploratory distribution of pathogenic variants in BRCA1 and BRCA2

Given that BRCA1 and BRCA2 harbored the highest number of pathogenic variants in the group of patients, an exploratory analysis was performed to visualize the distribution of these variants along the gene structures (Fig. 3A and B). Pathogenic variants were distributed across multiple regions of both genes, without evidence of clustering confined to a single functional domain. This exploratory visualization aimed to assess whether the pattern of pathogenic variants observed in this Chilean cohort aligned with or diverged from distributions commonly reported in other populations. While no distinct region-specific enrichment was evident, this analysis provides an initial framework for future comparative studies incorporating larger cohorts and population-specific data.

Fig. 3.

Exploratory distribution of pathogenic variants in BRCA1 and BRCA2. A Distribution of pathogenic and likely pathogenic variants along the BRCA1 gene structure. B Distribution of pathogenic and likely pathogenic variants along the BRCA2 gene structure. Variants are shown according to their relative position within each gene, illustrating the absence of region-specific clustering

Discussion

In this study, we characterized the germline variant landscape in Chilean patients with clinically suspected hereditary BC using a comprehensive exome-based approach. By extending the analysis beyond canonical BC predisposition genes, we identified a broad spectrum of pathogenic and likely pathogenic variants across multiple cancer susceptibility genes. Our findings highlight the substantial diagnostic gain achieved through expanded gene analysis and underscore the relevance of generating population-specific genomic data in underrepresented regions such as Latin America [7–9].

The use of WES enabled a broader and more flexible assessment of germline variation compared to predefined gene panels, particularly relevant in populations where the full spectrum of susceptibility genes is not yet fully characterized. A key finding of this study is the increased detection of pathogenic variants when expanding the analysis beyond the 20 NCCN-recommended BC genes. While canonical genes such as BRCA1 and BRCA2 accounted for a significant proportion of pathogenic findings, several clinically relevant variants were identified exclusively through the extended gene set. This observation supports previous reports demonstrating that multigene and exome-based approaches improve diagnostic yield compared with restricted gene panels, particularly in heterogeneous populations [4, 5, 28]. In settings where genetic data are limited, restricting testing to a small number of genes may underestimate hereditary cancer risk and contribute to inconclusive results.

The predominance of truncating variants observed in this group of patients is consistent with large international studies reporting loss-of-function alterations as the primary mechanism underlying hereditary cancer predisposition [3, 4]. However, the identification of pathogenic variants in multiple non-BRCA genes highlights a heterogeneous mutational landscape that may not be fully captured by traditional testing strategies. Similar observations have been reported in exome-based studies conducted in diverse and previously underrepresented populations, reinforcing the value of broad genomic analyses for uncovering population-specific variant spectra [10–12].

Beyond BRCA1 and BRCA2, pathogenic variants were identified in several additional cancer susceptibility genes, underscoring the genetic heterogeneity underlying hereditary BC. Although the clinical interpretation of non-BRCA variants remains challenging, particularly for genes with moderate or context-dependent penetrance, their inclusion contributes to reducing the proportion of genetically unexplained cases, as reported in previous multigene testing studies [29]. Importantly, the limited availability of population-specific genomic data in Latin American cohorts further complicates variant interpretation, emphasizing the need for expanded germline analyses combined with cautious and genetic counseling frameworks in underrepresented populations [30]. Importantly, the level of evidence supporting gene-disease associations varies across the extended set of cancer susceptibility genes included in this analysis. While some genes, such as BRCA1, BRCA2 and PALB2, have well-established roles in BC predisposition, others included in broader panels have more limited or context-dependent evidence. In this study, gene-disease validity was considered using ClinGen resources where available, and findings in genes with weaker associations were interpreted with caution. This distinction is particularly relevant when applying extended genomic analyses in underrepresented populations, where evidence supporting variant interpretation remains limited.

Although BRCA1 and BRCA2 remained the most frequently altered genes, exploratory visualization of pathogenic variants along these genes did not reveal clustering restricted to specific functional domains. Instead, variants were distributed across multiple regions, consistent with the diverse mutational patterns described in other populations [15–17]. While exploratory in nature, this analysis provides an initial reference for future studies aiming to assess whether regional or ancestry-related factors influence the distribution of pathogenic variants in high-risk genes.

The limited availability of comprehensive germline data in Chile and other Latin American countries poses a significant challenge for accurate variant interpretation and clinical decision-making. The Chilean population is characterized by a complex genetic admixture involving European, Indigenous, and African ancestries, which may influence the prevalence and spectrum of germline variants [18]. Underrepresentation of Latin American populations in large genomic reference databases, such as gnomAD, has been associated with higher rates of VUS and reduced confidence in variant classification [8, 9]. In this context, locally generated genomic data is essential to improve interpretation frameworks and support precision medicine initiatives. VUS identified in this study should be interpreted with caution, as their clinical relevance remains uncertain and requires further functional and population-based evidence.

From a clinical perspective, the identification of pathogenic variants in non-canonical cancer predisposition genes may have implications for genetic counselling, surveillance strategies, and cascade testing in affected families. Although the clinical actionability of some moderate-penetrance genes continues to be refined, their inclusion in comprehensive analyses provides valuable information for individualized risk assessment and multidisciplinary clinical management [4–6]. Careful interpretation remains essential, particularly in settings where population-specific data are still emerging.

Despite the results obtained, this study has some limitations to be considered. The retrospective and non-uniform nature of the available clinical data precluded a robust evaluation of the diagnostic yield associated with individual NCCN inclusion criteria. This represents an important area for future investigation, particularly in Latin American populations, where the applicability of guidelines developed in predominantly European ancestry populations remains to be fully assessed. The sample size, while representative of a clinically selected group of patients, limits the ability to draw definitive conclusions regarding population-level variant frequencies. In addition, ancestry markers were not evaluated, precluding formal analyses of genetic ancestry and its potential impact on variant distribution. Copy number variants (CNV) were not evaluated in this study because the exome capture design used for the initial sequencing batches was not optimized or validated for CNV detection. To ensure methodological consistency across the cohort, the analysis was restricted to single nucleotide variants and small insertions/deletions. Finally, functional validation of identified variants was beyond the scope of this study. Future investigations incorporating larger cohorts, ancestry-informed analyses, capture designs with analytical workflows tailored for CNV, and functional studies will be essential to further refine germline variant interpretation in this population.

Conclusions

In summary, this study provides a comprehensive overview of the germline variant spectrum in Chilean patients with suspected hereditary BC and demonstrates the added value of extended gene analysis beyond canonical predisposition genes. By contributing population-specific genomic data, these findings support ongoing efforts to improve variant interpretation and promote the equitable implementation of genomic medicine in underrepresented populations.

Supplementary Information

Below is the link to the electronic supplementary material.

Author contributions

S.M-P and R.F.R. conceived and designed the study. L.S. performed all laboratory procedures, including blood sample reception, DNA extraction, pre-analytical processing, and DNA storage. S.G-M, H.C.B.G., P.R.F.J., and A.S-B designed and implemented the bioinformatic pipeline for primary and secondary whole-exome sequencing data analysis. S.G-M and S.M-P performed data curation. A.E.S-P and I.A.S-M validated curated variants using commercial clinical variant interpretation platforms. M.J.D. and S.M-P carried out manual classification of clinically relevant genetic variants. S.G-M designed and prepared all figures included in the manuscript. S.M-P supervised the entire study. V.C.F., G.G.R., S.G-M and S.M-P contributed to writing the original draft of the manuscript. S.G-M and S.M-P contributed to the final review and editing of the manuscript. All authors have read and agreed to the published version of the manuscript.

Funding

This study was funded by ANID SIA 2022 grant number 85220068, FONDECYT de Iniciación 2024 grant number 11240411 and FIC Regional Atacama 2022 BIP 40057841 and Chamada CNPq/MCTI/FNDCT Nº 22/2024, Processo # 444940/2024-3. The funder played no role in study design, data collection, analysis and interpretation of data, or the writing of this manuscript.

Data availability

The data supporting the findings of this study are available within the article and its Supplementary Information. Due to the sensitive nature of germline genomic data and ethical restrictions associated with patient privacy, raw sequencing data are not publicly available. Additional information may be made available from the corresponding author upon reasonable request and subject to institutional and ethical approvals.

Code availability

Primary and secondary analyses were conducted using the open-source nf-core/sarek pipeline. Any additional custom scripts used for data filtering and visualization are available from the corresponding author upon reasonable request.

Declarations

Ethics approval and consent to participate

This study was conducted in accordance with the Helsinki Declaration and was approved by the Institutional Scientific Ethics Committee of Universidad Mayor (approval code: 0367:561/2023). All participants provided written informed consent prior to blood sample collection and genetic analysis.

Competing interests

The authors declare no competing interests.