MiHAs are often translated from noncanonical ORFs. HLA-I–restricted MiHAs were derived from annotated protein-coding reading frames or noncanonical ORFs in normal, alternative, or noncoding transcripts. (A) Of 159 MiHAs (new in orange, previously published in blue), 122 (76.7%) antigens are peptides created by missense SNPs leading to an AA change in canonical ORFs of protein-coding transcripts. This includes 1 antigen in an annotated ORF created by protein splicing. The other 37 (23.3%) antigens are cryptic peptides (light green) created by SNPs causing AA changes in noncanonical ORFs including missense or synonymous SNPs translated in out-of-frame ORFs, SNPs in upstream ORFs, alternative transcripts or lncRNA ORFs. (B) Four new MiHAs are derived from lncRNA ORFs. MiHAs (red boxes) with polymorphic residues (blue) are preceded by start codons. For LINC02427, a start codon is located in the 5′ upstream sequence (blue) of the complementary DNA (black).