| Neuropeptide discovery and characterization is challenging owning to the extensive posttranslational modifications, relatively short length of the bioactive peptides, limited sequence conservation, and highly regulated mechanism of release which limits the amount of material available for analysis. We present two approaches that will greatly facilitate neuropeptide discovery and characterization. One approach deals with the challenges of neuropeptide gene prediction, while the other provides a means to predict the bioactive neuropeptides from its precursor (prohormone).;Our genome-assisted neuropeptide discovery efforts employ bioinformatics and mass spectrometry to annotate neuropeptide genes from newly sequenced organisms. Specifically, we annotated 37, 25, and 43 prohormones, respectively, from the flour beetle (Tribolium castaneum), the purple sea urchin (Strongylocentrotus purpuratus), and the zebra finch (Taeniopygia guttata). Moreover, 98 purple sea urchin neuropeptides and 29 zebra finch neuropeptides were confirmed by mass spectrometry.;In addition, we developed a neuropeptide processing predictor trained on mammalian prohormones. This binary logistic regression method, which computes the cleavage probabilities at Lysine and Arginine residues, has 91.1% correct classification rate, 88% sensitivity, and 92.7% specificity. This and similar statistical models trained on insects and mollusk are made publicly available as a web-based program, NeuroPred, and can be found at http://neuroproteomics.scs.uiuc.edu/neuropred.html.;The methods employed for the discovery of neuropeptide genes and gene products in insect, echinoderms and chordates along with neuropeptide processing predictors present generalized and successful approaches for neuropeptide discovery in any organism. |
