Real-world comparison of genome- and exome-based diagnostic workflows in a pediatric Albanian rare disease cohort

Pediatric rare diseases are clinically and genetically heterogeneous, often resulting in delayed molecular diagnosis. In routine clinical practice, diagnostic outcomes reflect the combined performance of complete workflows that may differ in sample source, sequencing depth, laboratory procedures, bioinformatics, interpretation, and reporting. We compared two independent clinical diagnostic workflows applied in parallel to the same Albanian pediatric cohort. Seventy-two patients with suspected rare genetic disorders underwent genome sequencing (GS) from saliva or buccal swabs at approximately 30 × mean depth and exome sequencing (ES) from dried blood spots, with more than 98% of target bases covered at ≥20×. The ISO-accredited laboratories independently performed sequencing, analysis, interpretation, and reporting. We assessed report-level primary diagnostic yield, paired outcomes, variant classes, discordant findings, and laboratory-reported secondary/additional findings. Variants of uncertain significance (VUS) were excluded from yield. Raw data were not centrally reanalyzed, and sample types were not cross-tested. At least one workflow reported a pathogenic or likely pathogenic finding interpreted as a primary molecular diagnosis in 29/72 patients (40.3%; 95% CI, 28.9–52.5). GS reports contained primary diagnoses for 26/72 patients (36.1%; 95% CI, 25.1–48.3) and ES reports for 16/72 (22.2%; 95% CI, 13.3–33.6). Primary outcomes were concordant in 56 participants (13 positive and 43 negative) and discordant in 16 (13 GS-only and three ES-only). The GS laboratory also reported 13 secondary/additional variant-level rows in 11 patients (12 P/LP rows and one VUS row), including seven patients without a primary diagnosis. The workflows produced overlapping but non-identical clinical reports, and neither workflow was comprehensive. The observed differences cannot be attributed to sequencing technology alone because sample type, depth, laboratory, bioinformatics, prioritization, and reporting scope were not harmonized. Implementation decisions should integrate diagnostic contribution with cost, turnaround time, infrastructure, workforce, governance, and local feasibility.