An optimized WGS workflow for FFPE samples: Enabling high-confidence variant detection for MRD surveillance

This study aimed to define a robust and synergistic workflow for FFPE-based whole genome sequencing (WGS) by evaluating how different extraction and library preparation methods impact coverage uniformity and the reliable detection of complex, actionable variants.Despite representing the vast majority of tumor specimens, formalin-fixed paraffin-embedded (FFPE) samples present significant DNA quality issues that challenge variant calling accuracy and limit their potential for whole genome sequencing (WGS) in precision oncology. Optimizing DNA extraction and library preparation is therefore essential to maximize WGS data quality from FFPE samples, particularly for sensitive applications like tissue-informed Minimal Residual Disease (MRD) monitoring. In this study, we evaluated the impact of different workflow combinations on WGS data quality. We compared libraries from four GIAB FFPE control samples (HG002/3/4/5) prepared using two extraction methods and two library preparation kits. This work expands on previous findings by including bioinformatic analysis of complex variants, coverage uniformity, and clinically relevant variant calling. Performance was assessed via: 1) precision and accuracy of large insertion-deletions (INDELs >16 bp), 2) genome-wide coverage uniformity, and 3) SNP/INDEL variant calling in clinically actionable Tier 1A gene variants. Our analysis of large INDELs (>16 bp) confirmed trends observed in small variant data. The combined Covaris extraction and library prep workflow demonstrated significantly higher precision and accuracy compared to other combinations. Analysis of genome-wide coverage distribution showed that the workflow produced superior coverage uniformity, with a lower coefficient of variation (CV) and a higher percentage of the genome covered. Regarding the calling of clinically actionable Tier 1A variants, the Covaris workflow again achieved the highest F1, precision, and recall. These analyses demonstrate that a workflow optimized for both extraction and library preparation provides a robust solution for FFPE-based WGS. This approach ensures higher data quality, more uniform genome coverage, and more reliable detection of complex and clinically actionable variants. This high-fidelity variant calling directly addresses a central challenge in FFPE-based precision oncology, providing the confident variant detection essential for applications including tumor profiling and MRD surveillance.