遗传性癌症中的亲本来源感知基因组分析：仅使用先证者的血液样本识别家族中处于风险的一方

Background: Determining the parent of origin (PofO) of a variant in hereditary cancer guides counseling, risk management, recurrence risk assessment, and variant classification. In conditions involving genes with PofO effects, such as SDHD , SDHAF2 and MAX , this information can determine whether disease will manifest. Current approaches rely on family-based testing, yet uptake among eligible first-degree relatives remains low, with fewer than 30% undergoing testing, creating a major barrier in hereditary cancer.To address this gap, we developed Parent-of-Origin-Aware Genomic Analysis (POAga), a method that integrates chromosome-scale haplotyping with DNA methylation at differentially imprinted regions to assign PofO without parental data. To validate POAga, we applied it to individuals with hereditary cancer and known segregation of their pathogenic variants, and compared the predicted PofO with the established segregation to assess concordance and limitations. Methods: Blood samples are being collected from individuals with pathogenic variants in hereditary cancer genes, representing broad ranges of ages, ancestries, and cancer histories. Parental segregation was previously known or established through confirmatory testing. Predicted PofO is compared with true segregation to assess concordance. All samples undergo Strand-seq and long-read sequencing under an REB-approved protocol. Results: To date, 285 samples with 290 pathogenic variants have been analyzed across the following genes: BRCA2 (n=46), BRCA1 (n=42), MSH2 (n=34), SDHD (n=29), MLH1 (n=27), MSH6 (n=26), PMS2 (n=20), PALB2 (n=15), TP53 (n=14), ATM (n=14), CDH1 (n=9), CHEK2 (n=3), EPCAM (n=2), SDHAF2 (n=2), MUTYH (n=2), CDKN2A (n=1), POT1 (n=1), RAD51D (n=1), and SDHC (n=1). PofO was assigned for 250 variants, with 98.4% concordance (246/250). PofO could not be determined for 40 variants (13.8%, 40/290), mainly due to insufficient allele-specific methylation at imprinted regions or extended homozygosity that impeded phasing. Misassignments were rare and mainly due to stochastic phasing errors, unresolved inversions, or random allelic methylation at imprinted regions. Conclusion: POAga achieves clinical-grade accuracy in assigning PofO from a single blood sample in hereditary cancer. This directly addresses a major barrier in clinical genetics, particularly when parental samples are unavailable. For genes with PofO effects, this information can determine whether disease will manifest. By enabling reliable segregation without parental testing, POAga helps direct clinical efforts toward those truly at risk and improves the clinical interpretation of variants. Ongoing analyses will refine its performance and support its adoption as a transformative tool in hereditary cancer genomics.