Secondary Structure and Contact Guided Differential Evolution for Protein Structure Prediction.

Area of Science:

• Structural bioinformatics
• Computational biology
• Biophysics

Background:

• Ab initio protein tertiary structure prediction remains a significant challenge in structural bioinformatics.
• Leveraging predicted residue-residue contact and secondary structure information can improve prediction accuracy.

Purpose of the Study:

• To propose an improved differential evolution algorithm, SCDE, for enhanced ab initio protein structure prediction.
• To integrate secondary structure and residue-residue contact information into the prediction process.

Main Methods:

• Developed SCDE, incorporating two novel score models based on secondary structure and contact information.
• Designed secondary structure-based and contact-based selection strategies to guide conformational space search.
• Implemented a probability distribution function to balance the two selection strategies.

Main Results:

• Experimental validation on a benchmark dataset of 28 proteins and four CASP12 free modeling targets.
• Demonstrated the effectiveness and efficiency of the SCDE method in protein structure prediction.

Conclusions:

• The SCDE method significantly enhances the accuracy of ab initio protein tertiary structure prediction.
• SCDE offers an effective and efficient approach for tackling complex protein structure prediction problems.