Background: Formalin-fixed, paraffin-embedded (FFPE) tissues are a key resource for molecular oncology and retrospective genomic studies. However, formaldehyde fixation introduces characteristic DNA damage - primarily cytosine deamination, that can obscure true somatic variants and complicate further analyses, including clonality inference. Although multiple computational and enzymatic artifact-mitigation strategies exist, a systematic comparison across tools and experimental conditions has been lacking. Here, we evaluate five computational approaches (SOBDetector, Ideafix, MicroSEC, FFPolish, DeepOmics FFPE/FFPE-PLUS) along with the NEBNext® FFPE DNA Repair Mix v2, a multi-enzyme repair system, to assess their impact on data quality. Our analysis incorporated three datasets: 36 TCGA WES and 21 CGCI WGS cancers for which both FFPE and fresh-frozen samples were available, and a custom WES set containing fresh-frozen and formalin-treated samples processed with or without NEBNext repair. Variant calls were filtered and benchmarked against fresh-frozen references using som.py. Among computational strategies, DeepOmics FFPE and its WGS-optimized variant DeepOmics FFPE-PLUS showed the highest performance, providing an effective balance between artifact suppression and preservation of true somatic variants, particularly in WGS data. In parallel, enzymatic repair with NEBNext v2 produced the most consistent improvements in mutational profiles and concordance with fresh-frozen samples. We further assessed clonal architecture across fresh-frozen, FFPE, DeepOmics-processed FFPE, and NEBNext-repaired FFPE samples. While DeepOmics effectively reduced artifacts, its filtering distorted VAF distributions, compromising clonal reconstruction. NEBNext-treated samples most accurately recapitulated the clonal landscape of fresh-frozen material. Overall, enzymatic repair preserved biological signal more effectively, offering more reliable clonality analyses of FFPE-derived data. Funding: NCN Grant: 2021/41/B/NZ2/04134; 02/040/BKM25/1074