Transmembrane protein alignment and fold recognition based on predicted topology.

Area of Science:

• Structural Biology
• Bioinformatics
• Computational Biology

Background:

• Transmembrane proteins (TMPs) are crucial for biological processes and drug development.
• Predicting TMP structures remains challenging due to their unique properties compared to soluble proteins.
• Existing fold recognition methods are less developed for TMPs than for soluble proteins.

Purpose of the Study:

• To develop a novel method for Transmembrane Protein Fold Recognition (TMPFR).
• To improve the accuracy of TMP fold recognition and sequence-to-structure alignment.
• To address the limitations of current prediction tools for TMPs.

Main Methods:

• Developed TMFR, a novel TMP fold recognition method based on sequence-to-structure pairwise alignment.
• Utilized topology-based features, sequence profiles, and solvent accessibility in alignment.
• Incorporated a gap penalty dependent on predicted topology structure segments.
• Trained scoring function parameters separately for alpha-helical (αTMP) and beta-strand (βTMP) proteins.

Main Results:

• TMFR achieved 10% higher accuracy for αTMPs and 9% higher accuracy for βTMPs compared to HHalign.
• The raw score from TMFR is negatively correlated with structure similarity, indicating effective fold recognition.
• Demonstrated TMFR as an effective TMP-specific fold recognition and alignment method.

Conclusions:

• TMFR offers a significant advancement in TMP structure prediction.
• The method's specificity for TMPs leads to improved accuracy.
• This work provides a valuable tool for studying TMP structures and facilitating pharmaceutical development.