SRA

Although ribosome profiling and translation initiation sequencing (TI-seq) analyses identifies a large number of non-canonical initiation codons, precise detection of translation initiation sites (TIS) remains a challenge, mainly due to the experimental artifacts of those analyses. Here we develop a method termed TISCA (TIS detection by translation Complex Analysis) for accurate identification of the TISs. TISCA shows higher reliability for TIS detection than other existing tools, and identify a significant number of near-cognate codons adjacent to Kozak-like sequence contexts. Proteomics analysis reveals that the majority of the NH2-terminus of proteins derived from the near-cognate initiation codons is methionine. Although eIF2, eIF2A, and eIF2D are known to contribute to translation initiation at near-cognate codons, most non-canonical initiation events are eIF2-dependent, consistent with the initial amino acid being methionine. The comprehensive identification of TISs facilitates the analysis of non-canonical initiation mechanisms.