Detection of Pathogenic Variants With Germline Genetic Testing Using Deep Learning vs Standard Methods in Patients With Prostate Cancer and Melanoma

AlDubayan, Saud H.; Conway, Jake R.; Camp, Sabrina Y.; Witkowski, Leora; Kofman, Eric; Reardon, Brendan; Han, Seung-Hun; Moore, N.; Elmarakeby, Haitham; Salari, Keyan; Choudhry, Hani; Al-Rubaish, Abdullah M.; Al-Sulaiman, Abdulsalam A.; Al-Ali, Amein K.; Taylor-Weiner, Amaro; Allen, Eliezer M. Van

doi:10.1001/jama.2020.20457

Cited by 39 publications

(22 citation statements)

References 32 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…By analyzing each variant in the context of specimen purity, we eliminate the need for ad hoc VAF criteria, 14 or complex analyses of raw sequencing data. 36 …”

Section: Discussionmentioning

confidence: 99%

Germline Testing Data Validate Inferences of Mutational Status for Variants Detected From Tumor-Only Sequencing

Jalloul

Gomy

Stokes

et al. 2021

JCO Precision Oncology

View full text Add to dashboard Cite

PURPOSE Pathogenic germline variants (PGVs) in cancer susceptibility genes are usually identified through germline testing of DNA from blood or saliva: their detection can affect treatment options and potential risk-reduction strategies for patient relatives. PGV can also be identified in tumor sequencing assays, which, when performed without patient-matched normal specimens, render determination of variants' germline or somatic origin critical. METHODS Tumor-only sequencing data from 1,608 patients were retrospectively analyzed to infer germline versus somatic status of variants using an information-theoretic, gene-independent approach. Loss of heterozygosity was also determined. Predicted mutational models were compared with clinical germline testing results. Statistical measures were computed to evaluate performance. RESULTS Tumor-only sequencing detected 3,988 variants across 70 cancer susceptibility genes for which germline testing data were available. We imputed germline versus somatic status for > 75% of all detected variants, with a sensitivity of 65%, specificity of 88%, and overall accuracy of 86% for pathogenic variants. False omission rate was 3%, signifying minimal error in misclassifying true PGV. A higher portion of PGV in known hereditary tumor suppressors were found to be retained with loss of heterozygosity in the tumor specimens (72%) compared with variants of uncertain significance (58%). CONCLUSION Analyzing tumor-only data in the context of specimens' tumor cell content allows precise, systematic exclusion of somatic variants and suggests a balance between type 1 and 2 errors for identification of patients with candidate PGV for standard germline testing. Although technical or systematic errors in measuring variant allele frequency could result in incorrect inference, misestimation of specimen purity could result in inferring somatic variants as germline in somatically mutated tumor suppressor genes. A user-friendly bioinformatics application facilitates objective analysis of tumor-only data in clinical settings.

show abstract

“…By analyzing each variant in the context of specimen purity, we eliminate the need for ad hoc VAF criteria, 14 or complex analyses of raw sequencing data. 36 …”

Section: Discussionmentioning

confidence: 99%

Germline Testing Data Validate Inferences of Mutational Status for Variants Detected From Tumor-Only Sequencing

Jalloul

Gomy

Stokes

et al. 2021

JCO Precision Oncology

View full text Add to dashboard Cite

show abstract

“…In this work, we build and assess DeepTrio, a deep learning-based variant caller for parent-child trios. We start from the code base of DeepVariant, a germline caller which won multiple awards in the PrecisionFDA Truth Challenge V2 22 , noted for high accuracy on genomes and exomes 24 , and shown to increase detection rate of pathogenic germline variants 25 .…”

Section: Introductionmentioning

confidence: 99%

DeepTrio: Variant Calling in Families Using Deep Learning

Kolesnikov

Goel

Nattestad

et al. 2021

Preprint

View full text Add to dashboard Cite

Every human inherits one copy of the genome from their mother and another from their father. Parental inheritance helps us understand the transmission of traits and genetic diseases, which often involve de novo variants and rare recessive alleles. Here we present DeepTrio, which learns to analyze child-mother-father trios from the joint sequence information, without explicit encoding of inheritance priors. DeepTrio learns how to weigh sequencing error, mapping error, and de novo rates and genome context directly from the sequence data. DeepTrio has higher accuracy on both Illumina and PacBio HiFi data when compared to DeepVariant. Improvements are especially pronounced at lower coverages (with 20x DeepTrio roughly equivalent to 30x DeepVariant). As DeepTrio learns directly from data, we also demonstrate extensions to exome calling solely by changing the training data. DeepTrio includes pre-trained models for Illumina WGS, Illumina exome, and PacBio HiFi.

show abstract

“…total depth, focal ploidy, and VAF) can select the most likely germline versus somatic mutational status and assess evidence for loss of heterozygosity. By analyzing each variant in the context of specimen purity, we eliminate the need for ad hoc VAF criteria [14], or complex analyses of raw sequencing data [35]. We validated our approach using available germline testing results from 1,608 cancer patients.…”

Section: Discussionmentioning

confidence: 99%

Germline testing data validate inferences of mutational status for variants detected from tumor-only sequencing

Jalloul

Gomy

Stokes

et al. 2021

Preprint

View full text Add to dashboard Cite

Structured AbstractBackgroundPathogenic germline variants (PGV) in cancer susceptibility genes are usually identified in cancer patients through germline testing of DNA from blood or saliva: their detection can impact patient treatment options and potential risk reduction strategies for relatives. PGV can also be identified, in tumor sequencing assays, often performed without matched normal specimens. It is then critical to determine whether detected variants are somatic or germline. Here, we evaluate the clinical utility of computational inference of mutational status in tumor-only sequencing compared to germline testing results.Patients and MethodsTumor-only sequencing data from 1,608 patients were retrospectively analyzed to infer germline-versus-somatic status of variants using an information-theoretic, gene-independent approach. Loss of heterozygosity (LOH) was also determined. The predicted mutational models were compared to clinical germline testing results. Statistical measures were computed to evaluate performance.ResultsTumor-only sequencing detected 3,988 variants across 70 cancer susceptibility genes for which germline testing data were available. Our analysis imputed germline-versus-somatic status for >75% of all detected variants, with a sensitivity of 65%, specificity of 88%, and overall accuracy of 86% for pathogenic variants. False omission rate was 3%, signifying minimal error in misclassifying true PGV. A higher portion of PGV in known hereditary tumor suppressors were found to be retained with LOH in the tumor specimens (72%) compared to variants of uncertain significance (58%).ConclusionsTumor-only sequencing provides sufficient power to distinguish germline and somatic variants and infer LOH. Although accurate detection of PGV from tumor-only data is possible, analyzing sequencing data in the context of specimens’ tumor cell content allows systematic exclusion of somatic variants, and suggests a balance between type 1 and 2 errors for identification of patients with candidate PGV for standard germline testing. Our approach, implemented in a user-friendly bioinformatics application, facilities objective analysis of tumor-only data in clinical settings.HighlightsMost pathogenic germline variants in cancer predisposition genes can be identified by analyzing tumor-only sequencing data.Information-theoretic gene-independent analysis of common sequencing data accurately infers germline vs. somatic status.A reasonable statistical balance can be established between sensitivity and specificity demonstrating clinical utility.Pathogenic germline variants are more often detected with loss of heterozygosity vs. germline variants of uncertain significance.

show abstract

Detection of Pathogenic Variants With Germline Genetic Testing Using Deep Learning vs Standard Methods in Patients With Prostate Cancer and Melanoma

Cited by 39 publications

References 32 publications

Germline Testing Data Validate Inferences of Mutational Status for Variants Detected From Tumor-Only Sequencing

Germline Testing Data Validate Inferences of Mutational Status for Variants Detected From Tumor-Only Sequencing

DeepTrio: Variant Calling in Families Using Deep Learning

Germline testing data validate inferences of mutational status for variants detected from tumor-only sequencing

Contact Info

Product

Resources

About