A python analytical pipeline to identify prohormone precursors and predict prohormone cleavage sites

Area of Science:

• Neuroscience
• Bioinformatics
• Genomics

Background:

• Neuropeptides and hormones are crucial signaling molecules in the central nervous system.
• Experimental characterization of neuropeptides is challenging due to complex precursor processing.
• Bioinformatic approaches are needed to streamline neuropeptide identification.

Purpose of the Study:

• To develop a flexible Python-based bioinformatics pipeline for identifying neuropeptide precursors and predicting cleavage sites.
• To create a user-centered web application (NeuroPred) for the neuroscience community.
• To demonstrate the utility of Python in comprehensive neuropeptide prediction.

Main Methods:

• Utilized Python to develop bioinformatics pipeline components for precursor identification from genomic data.
• Employed support vector machines (SVMs) to predict neuropeptide cleavage sites.
• Compared SVM predictions for rhesus macaque with human sequence homology.
• Developed the NeuroPred web application for cleavage site prediction and analysis.

Main Results:

• Identified 75 neuropeptide precursors in the rhesus genome.
• Achieved over 97% cleavage prediction accuracy using SVMs for both human and rhesus datasets.
• NeuroPred provides predictions, statistics, post-translational modifications, and peptide molecular mass.

Conclusions:

• Python is a powerful and flexible language for developing comprehensive bioinformatics tools.
• The developed pipeline and NeuroPred effectively predict neuropeptides from genomic data.
• This approach facilitates the identification of bioactive peptides, advancing neuroscience research.