Table 1.
Sensitivity and specificity analysis of BayPR
| Gene | Variant Interpretation | No. Variants Scored | BayPR Score Thresholda (%) | Positive n (%) | Negative n (%) | Sensitivity (95% CI) | Specificity (95% CI) | BayPR Score Thresholda | Positive n (%) | Negative n (%) | Sensitivity (95% CI) | Specificity (95% CI) |
|---|---|---|---|---|---|---|---|---|---|---|---|---|
| CFTR | Pathogenic/likely pathogenic | 296 | 90 | 284 (95.9) | 12 (4.1) | 95.9 (93.0–97.9) | 94.1 (71.3–99.9) | 80 | 286 (96.6) | 10 (3.4) | 96.6 (93.9–98.4) | 94.1 (71.3–99.9) |
| Benign/likely benign | 17 | 10 | 1 (5.9) | 16 (94.1) | 20 | 1 (5.9) | 16 (94.1) | |||||
| PAH | Pathogenic/likely pathogenic | 393 | 90 | 319 (81.2) | 74 (18.8) | - | - | 80 | 385 (98.0) | 8 (2.0) | - | - |
| ABCA3 | Pathogenic/likely pathogenic | 225 | 90 | 224 (99.6) | 1 (0.4) | 99.6 (97.5–100.0) | 0.0 (0.0–52.2) | 80 | 225 (100.0) | 0 (0.0) | 100.0 (98.4–100.0) | 100.0 (47.8–100.0) |
| Benign/likely benign | 5 | 10 | 0 (0.0) | 5 (100.0) | 20 | 0 (0.0) | 5 (100.0) | |||||
| FBN1 | Pathogenic/likely pathogenic | 178 | 90 | 178 (100.0) | 0 (0.0) | - | - | 80 | 178 (100.0) | 0 (0.0) | - | - |
| TGFBR1 | Pathogenic/likely pathogenic | 23 | 90 | 2 (8.7) | 21 (91.3) | 80 | 21 (91.3) | 2 (8.7) | ||||
| TGFBR2 | Pathogenic/likely pathogenic | 55 | 90 | 2 (3.6) | 53 (96.4) | - | - | 80 | 54 (98.2) | 1 (1.8) | - | - |
| ABCD1 | Pathogenic/likely pathogenic | 61 | 90 | 61 (100.0) | 0 (0.0) | 100.0 (94.1–100.0) | 100.0 (2.5–100.0) | 80 | 61 (100.0) | 0 (0.0) | 100.0 (94.1–100.0) | 100.0 (2.5–100.0) |
| Benign/likely benign | 1 | 10 | 0 (100.0) | 1 (0.0) | 20 | 0 (100.0) | 1 (0.0) | |||||
| TAFAZZIN | Pathogenic/likely pathogenic | 143 | 90 | 41 (28.7) | 102 (71.3) | - | - | 80 | 140 (97.9) | 3 (2.1) | - | - |
BayPR, Bayesian prevalence ratio; CI, confidence interval.
a
Score thresholds were set to determine whether a variant had a positive or negative result from the BayPR for the purposes of sensitivity and specificity analysis. For pathogenic/likely pathogenic variants, BayPR scores >90% or >80% were considered true positives. For benign/likely benign variants, Bayesian scores <10% or <20% were considered true negatives. Pathogenic/likely pathogenic and benign/likely benign determinations were made by expert review. Only genes with both pathogenic/likely pathogenic and benign/likely benign underwent sensitivity and specificity analysis.