Venom peptides have proven to be a very rich source for novel drugs leads and invaluable research tools, helping to dissect physiological functions of many human receptors and elucidating the biological mechanisms underlying diseases. Neuropeptides are present throughout the animal kingdom and several have been found in animal venoms including oxytocin and insulin. Neuropeptides are implicated in many diseases and the identification of novel neuropeptide sequences in animal venoms will not only broaden our understanding of animal venom composition, but could lead to novel research tools as well as drug leads.
Here we present a novel bioinformatics workflow that searches in-house as well as publicly available animal venom transcriptomes for novel neuropeptide sequences. We used profile Hidden Markov Models to search venom transcriptomes including amongst others snakes, spiders, centipedes and cone snails. Integration of tools such as SignalP, NeuroPred and CD-HIT allow for deeper mining and accurate downstream annotation.
A systematic search for neuropeptides in animal venoms has not been done before. The initial workflow identified novel oxytocin and insulin sequences from 18 cone snail venom transcriptomes, providing the proof-of-concept. Automation and upscaling of that workflow has allowed identification of neuropeptide sequences homologous to human sequences from centipedes, spiders and snakes, in addition to cone snails. The current bioinformatics workflow provides a very cost-effective animal venom neuropeptide discovery tool.