We aim to apply artificial intelligence (AI) models in combination with experimental assays to unveil the roles of the newly discovered non-canonical proteome in the cellular physiology of G protein-coupled receptors (GPCRs). Genes have historically been thought to encode a single reference protein and its spliced isoforms, but recent technological advances (Ribo-Seq and proteogenomics) have highlighted thousands of non-canonical coding sequences (CDS)[1-2]. These non-canonical sequences are found in non-coding RNAs (ncRNAs), in the untranslated regions (UTRs) of mRNAs, or overlapping the canonical CDS in a different frame. Genomes can thereby generate many proteins aside from the currently annotated proteins. These additional proteins are termed alternative proteins or AltProts[1,3]. An ever-increasing number of studies highlight the biological roles of AltProts, notably their ability to bind to, and modulate, the functions of the canonical protein product of the gene[1,4-9]. This alternative proteome represents a large reservoir of uncharacterized proteins and hitherto unexplored avenues of fundamental and clinical research. However, there is no standardized annotation of AltProts and as they are relatively small, the risk of random CDS (false positive) is high. We need a data-driven and reliable approach to identify translated CDS and prioritize AltProts for functional characterization. Recent successes of deep learning models in sequence segmentation and classification bodes well for its applicability to the field of AltProts. Here, we suggest deep learning models can guide AltProt prioritization as well as structural and functional characterization to unveil their roles in the regulation of cell signalling. Our efforts will focus on GPCR signalling pathways as they control cellular physiology and are major targets for therapeutic drugs[10-11]. Identifying AltProts that are novel regulators of these pathways will transform our understanding of cell biology and generate actionable knowledge for the pharmacological industry. This interdisciplinary project links AI, bioinformatics, genetics, proteomics, structural biology, pharmacology and cell biology. It aims to use AI to identify and prioritize AltProts involved in regulation of GPCR signalling in human cells, and to experimentally validate predictions.