This study addresses the issue of missing heritability, where the genetic variance identified through significant markers is less than the actual genetic variance estimated through family. Using a newly developed multi-omics analysis workflow on the tomato graph pangenome, the study explores the missing heritability from three aspects: incomplete linkage disequilibrium, allelic heterogeneity, and locus heterogeneity. Furthermore, the findings indicate that recovering missing heritability can enhance molecular marker-assisted selection and genomic selection approaches for tomato flavor breeding.