Comparison study on statistical features of predicted secondary structures for protein structural class prediction:...

Area of Science:

• Bioinformatics
• Computational Biology
• Structural Biology

Background:

• Protein structural class prediction is crucial in bioinformatics.
• Existing methods utilize content-based statistical features of secondary structural elements (CBF-PSSEs).
• The positional distribution of successive elements in predicted secondary structures has not been fully explored.

Purpose of the Study:

• To introduce and evaluate novel position-based features of predicted secondary structural elements (PBF-PSSEs).
• To systematically compare the predictive power of PBF-PSSEs against established CBF-PSSEs.
• To investigate the combined performance of PBF-PSSEs and CBF-PSSEs for improved protein classification.

Main Methods:

• Development and implementation of PBF-PSSEs, focusing on the positional distribution of secondary structural elements.
• Quantitative assessment of PBF-PSSEs and CBF-PSSEs individually.
• Creation of a combined feature set (PBF11CBF-PSSE) integrating both feature types.

Main Results:

• PBF-PSSEs demonstrate significant predictive ability for protein structural class.
• Combining PBF-PSSEs with CBF-PSSEs leads to substantial improvements in prediction accuracy.
• A novel combined feature set, PBF11CBF-PSSE, was constructed and evaluated.

Conclusions:

• PBF-PSSEs offer a promising avenue for enhancing protein structural class prediction.
• The integration of positional and content-based features yields complementary and significant contributions.
• Optimizing the parameter 'k' for PBF-PSSEs is critical for maximizing performance.