Abstract 2041: Research strategies for rare germline variants and cancer study

Abstract Exploration of rare germline variants in cancer is the fields of ongoing research. So, we would like to suggest several considerations for the rare pathogenic germline variants studies. We confirmed the pathogenicity of each variant using ClinVar (https://www.ncbi.nlm.nih.gov/clinvar). In addition, we utilized in silico prediction tools such as Polyphen and SIFT. Even for the same variant, the report of the variant may vary depending on the database used. We conducted rare germline variants screening using ICGC (International Cancer Genome Consortium) PCAWG (Pan-Cancer Analysis of Whole Genomes) data. we filtered allele frequencies using the gnomAD (v2.1.1) exome database. We utilized several ClinVar versions to check the clinical significance of germline variants. Next, we have investigated some in silico tools that predict ClinVar pathogenic and benign variants well. We observed inconsistent variant reports across versions of ClinVar. Although the information on variants may change over time, there have been inconsistent changes in the clinical significance of variants reported as P (Pathogenic) or LP (Likely pathogenic). The clinical significance of the variants reported in the previous version was P or LP, but those in the updated version were changed to “Conflicting interpretations of pathogenicity”. It was confirmed that the variant report changes when at least one “Uncertain Significance” review is added. Among various prediction tools, we found that the predictive power of a tool called “MetaRNN” was useful. Compared with several tools, we confirmed that MetaRNN’s AUC (Area Under the ROC Curve) value was 0.985, which is the best performance. But, it is not enough to judge the pathogenicity of a variant only by the prediction result of the tool. And, prediction tools are also known to be more efficient in predicting missense variants than others. Therefore, it is necessary to examine the pathogenicity of germline variants using both the prediction tool and the clinical significance of ClinVar. In summary, for rare germline variants in cancer studies, it is necessary to observe variants compared to the previous versions of ClinVar. Moreover, if prediction tools are appropriately used, researchers will increase the possibility of finding a variant or gene with more crucial clinical significance to reveal the association between the rare germline variants and cancer development. Citation Format: Seokhyeon Kim, Youngil Koh, Sung-Soo Yoon. Research strategies for rare germline variants and cancer study [abstract]. In: Proceedings of the American Association for Cancer Research Annual Meeting 2023; Part 1 (Regular and Invited Abstracts); 2023 Apr 14-19; Orlando, FL. Philadelphia (PA): AACR; Cancer Res 2023;83(7_Suppl):Abstract nr 2041.