PONGO

PONGO integrates TMHMM2.0, MEMSAT, PRODIV, ENSEMBLE1.0 and SPEP (SPEPlip) to predict membrane protein topology and signal peptides for analysis of protein sorting and genome annotation.

Key Features:

• Integration of Multiple Predictors: Integrates TMHMM2.0, MEMSAT, PRODIV and ENSEMBLE1.0 to produce comparative topology predictions for membrane proteins.
• Comparative prediction evaluation: Enables simultaneous evaluation and comparison of multiple topology predictions for a single sequence.
• Signal Peptide Prediction: Incorporates SPEP (SPEPlip), a neural network trained on experimentally derived eukaryotic and prokaryotic signal peptides, to identify sorting signals and predict cleavage sites with low reported false positive and false negative rates.
• Organism scope: Supports analysis of putative membrane proteins from diverse organisms while maintaining consistent sequence profiling across inputs.

Scientific Applications:

• Membrane protein topology and structure: Comparative topology predictions inform inference of membrane protein topology and potential structural features.
• Genome annotation: Provides topology and signal peptide annotations useful for annotating genes in genomic datasets.
• Protein sorting and signal peptide analysis: Identifies sorting signals and cleavage sites to study protein targeting and sorting mechanisms.

Methodology:

Leverages distributed annotation systems (DAS) developed by BIOSAPIENS and processes input sequences with TMHMM2.0, MEMSAT, PRODIV, ENSEMBLE1.0 and SPEP (SPEPlip) to generate topology predictions and signal peptide annotations.