The rise of high-throughput RNA sequencing (RNA-seq) and de novo transcriptome assembly has had a transformative impact on how we identify and study genes in the phototransduction cascade of non-model organisms. But the advantage provided by the nearly automated annotation of RNA-seq transcriptomes may at the same time hinder the possibility for gene discovery and the discovery of new gene functions. For example， standard functional annotation based on domain homology to known protein families can only confirm group membership， not identify the emergence of new biochemical function. In this study， we show the importance of developing a strategy that circumvents the limitations of semiautomated annotation and apply this workflow to photosensitivity as a means to discover non-opsin photoreceptors. We hypothesize that non-opsin G-protein-coupled receptor (GPCR) proteins may have chromophore-binding lysines in locations that differ from opsin. Here， we provide the first case study describing non-opsin light-sensitive GPCRs based on tissue-specific RNA-seq data of the common bay scallop Argopecten irradians (Lamarck， 1819). Using a combination of sequence analysis and three-dimensional protein modeling， we identified two candidate proteins. We tested their photochemical properties and provide evidence showing that these two proteins incorporate 11-cis and/or all-trans retinal and react to light photochemically. Based on this case study， we demonstrate that there is potential for the discovery of new light-sensitive GPCRs， and we have developed a workflow that starts from RNA-seq assemblies to the discovery of new non-opsin， GPCR-based photopigments.