Abstract
Formalin-Fixed Paraffin-Embedded (FFPE) samples are routinely used for genome sequencing as companion diagnostics for targeted therapies for cancer patients. While targeted sequencing of FFPE tissue usually provides accurate assessment of hotspot mutations, whole exome or genome sequencing data are loaded with artefacts. Standardized and optimized methods are needed to overcome these obstacles and ensure the reliability of clinical exome/genome sequencing data. We evaluated the impact of the choice of DNA extraction kits, fragmentation methods, and input amount on data quality. We also evaluated the ability of repair enzymes and bioinformatics algorithms to remove FFPE induced artefacts and mutation signatures. This study discovered that employing the ReliaPrep extraction kit significantly enhances DNA yield from FFPE samples. Notably, the application of DNA repair enzymes emerged as a pivotal factor in suppressing artefacts, thereby enhancing the reliability of somatic mutation identification. Ultrasonication-based fragmentation proves more advantageous in reducing artefacts and improving the reproducibility. Additionally, bioinformatic algorithms efficiently eradicated any remaining artefacts, leading to improved reproducibility and accurate mutation signature identification. These findings provide a comprehensive methodology for handling FFPE samples in clinical research, ensuring the reliability and relevance of genomic data for personalized medicine and cancer treatment strategies.