T1SEstacker: A tri-layer stacking model effectively predicts bacterial type 1 secreted proteins based on C-terminal non-RTX-motif sequence features

Abstract

The proteins secreted through type 1 secretion systems often play important roles in pathogenicity of various gram-negative bacteria. However, the type 1 secretion mechanism remains unknown. In this research, we observed the sequence features of RTX proteins, a major class of type 1 secreted substrates. We found striking non-RTX-motif amino acid composition patterns at the C-termini, most typically exemplified by the enriched [FLI][VAI] at the most C-terminal two positions. Machine-learning models, including deep-learning models, were trained using these sequence-based non-RTX-motif features, and further combined into a tri-layer stacking model, T1SEstacker, which predicted the RTX proteins accurately, with a 5-fold cross-validated sensitivity of ~0.89 at the specificity of ~0.94. Besides substrates with RTX motifs, T1SEstacker can also well distinguish non-RTX-motif type 1 secreted proteins, further suggesting their potential existence of common secretion signals. In summary, we made comprehensive sequence analysis on the type 1 secreted RTX proteins, identified common sequence-based features at the C-termini, and developed a stacking model that can predict type 1 secreted proteins accurately.