We commonly call genes that share common ancestors homolog. But what about homolog of two kinds of neuropeptides? How can we find them?
I find the following case from the HHMI.
"These neurons express a neuropeptide, tachykinin, whose vertebrate homolog substance P has been implicated in the control of aggression in several mammalian systems."
http://www.hhmi.org/research/neural-circuits-innate-emotional-behaviors-flies-and-mice
There are various prohormone cleavage prediction bioinformatics tools that greatly aid annotation of neuropeptide genes. I respectfully recommend a few following articles describing some of these approaches. Those are tried and true at my institution and have lead to de novo neuropeptide discovery in honey bee, song bird, and planaria among other species.
http://www.ncbi.nlm.nih.gov/pmc/articles/PMC2610252/
http://www.ncbi.nlm.nih.gov/pmc/articles/PMC2698874/
You should remember to differentiate between the neuropeptide precursor proteins (NPPs), and the neuropeptides/peptide products themselves.
As for prediction of the NPs and NPPs, in addition to neuropred, as mentioned, for predicting NP products; i'd also suggest looking at:
http://neuropid.cs.huji.ac.il/
for predicting NPPs themselves.
http://neuropid.cs.huji.ac.il/
http://neuroproteomics.scs.illinois.edu/cgi-bin/neuropred.py
Clarification to my previous post: Included links to the original research articles in my previous post refer to the bioinformatics pipeline based on Python and machine learning for annotation of neuropeptide prohormones from unannotated genomic info. Having prohormones identified, obviously, helps neuropeptide identification as a follow-up step and here Neuropred come handy.