Comprehensive Gene Panel Testing for Hearing Loss in Children: Understanding Factors Influencing Diagnostic Yield.

OBJECTIVE: To evaluate factors influencing the diagnostic yield of comprehensive gene panel testing (CGPT) for hearing loss (HL) in children and to understand the characteristics of undiagnosed probands. STUDY DESIGN: This was a retrospective cohort study of 474 probands with childhood-onset HL who underwent CGPT between 2016 and 2020 at a single center. Main outcomes and measures included the association between clinical variables and diagnostic yield and the genetic and clinical characteristics of undiagnosed probands. RESULTS: The overall diagnostic yield was 44% (209/474) with causative variants involving 41 genes. While the diagnostic yield was high in the probands with congenital, bilateral, and severe HL, it was low in those with unilateral, noncongenital, or mild HL; cochlear nerve deficiency; preterm birth; neonatal intensive care unit admittance; certain ancestry; and developmental delay. Follow-up studies on 49 probands with initially inconclusive CGPT results changed the diagnostic status to likely positive or negative outcomes in 39 of them (80%). Reflex to exome sequencing on 128 undiagnosed probands by CGPT revealed diagnostic findings in 8 individuals, 5 of whom had developmental delays. The remaining 255 probands were undiagnosed, with 173 (173/255) having only a single variant in the gene(s) associated with autosomal recessive HL and 28% (48/173) having a matched phenotype. CONCLUSION: CGPT efficiently identifies the genetic etiologies of HL in children. CGPT-undiagnosed probands may benefit from follow-up studies or expanded testing.

Using Machine Learning to Identify True Somatic Variants from Next-Generation Sequencing.

BACKGROUND: Molecular profiling has become essential for tumor risk stratification and treatment selection. However, cancer genome complexity and technical artifacts make identification of real variants a challenge. Currently, clinical laboratories rely on manual screening, which is costly, subjective, and not scalable. We present a machine learning-based method to distinguish artifacts from bona fide single-nucleotide variants (SNVs) detected by next-generation sequencing from nonformalin-fixed paraffin-embedded tumor specimens. METHODS: A cohort of 11278 SNVs identified through clinical sequencing of tumor specimens was collected and divided into training, validation, and test sets. Each SNV was manually inspected and labeled as either real or artifact as part of clinical laboratory workflow. A 3-class (real, artifact, and uncertain) model was developed on the training set, fine-tuned with the validation set, and then evaluated on the test set. Prediction intervals reflecting the certainty of the classifications were derived during the process to label "uncertain" variants. RESULTS: The optimized classifier demonstrated 100% specificity and 97% sensitivity over 5587 SNVs of the test set. Overall, 1252 of 1341 true-positive variants were identified as real, 4143 of 4246 false-positive calls were deemed artifacts, whereas only 192 (3.4%) SNVs were labeled as "uncertain," with zero misclassification between the true positives and artifacts in the test set. CONCLUSIONS: We presented a computational classifier to identify variant artifacts detected from tumor sequencing. Overall, 96.6% of the SNVs received definitive labels and thus were exempt from manual review. This framework could improve quality and efficiency of the variant review process in clinical laboratories.