MetaTIS: a tool to predict cognate and near-cognate translation initiation sites in human

Ribosomes typically commence translation at a methionine-encoding AUG codon flanked by a so-called Kozak region, a short nucleic acid motif that serves as an initiation site in humans. Though, the characteristic AUG start codon of an mRNA is not always effective in initiating translation. Near-cognate codons differing from AUG by one nucleotide may also be recognized as start sites. Several types of ribosomal profiling techniques have been developed that elucidate active translation initiation sites (TIS) that enable training of computational models to predict both cognate and near-cognate TIS using mRNA sequence features. Here, a meta-model termed MetaTIS was implemented by combining outputs of genomic and protein language models fine-tuned on Ensembl annotations of transcripts and five different TIS datasets. The model proficiently differentiates between spurious and true TIS in four distinct test sets, for both AUG and non-AUG instances. Most important for translation initiation based on one of the base model outputs was the Kozak sequence context and a region further upstream in the 5'UTR [-12, -10]. MetaTIS is available as a webserver at https://service2.bioinformatik.uni-saarland.de/metatis/, a tool that accurately predicts TIS for AUG and nine near-cognate start codons.