Protein structure prediction constrained by solution X-ray scattering data and structural homology identification.

Area of Science:

• Structural biology
• Computational biology
• Biophysics

Background:

• Protein structure prediction is crucial for understanding biological function.
• Experimental data, like Small Angle X-ray Scattering (SAXS), can guide computational predictions.
• SAXS data provides distance distributions but lacks residue-specific information, posing a challenge for structure prediction.

Purpose of the Study:

• To explore the utility of distance constraints from SAXS measurements in filtering candidate protein structures.
• To assess the effectiveness of SAXS data in improving the reliability of ab initio protein structure prediction.
• To investigate the potential of SAXS-guided prediction for identifying structural homologies and inferring protein function.

Main Methods:

• Extended the Levitt-Hinds coarse-grained approach for ab initio protein structure prediction.
• Generated candidate C(alpha) backbones using this extended method.
• Implemented a SAXS-based filter to refine the set of predicted structures.
• Evaluated predicted structures using structural homology searches against the Dali domain library.

Main Results:

• SAXS filtering effectively purified native structure candidates, significantly improving prediction reliability.
• Despite limitations in local detail and C(alpha) backbone accuracy, the approach extracted useful fold classification information.
• Structural homology searches against the Dali domain library yielded encouraging results.

Conclusions:

• SAXS data, despite its inherent limitations, can be successfully integrated into protein structure prediction pipelines.
• This SAXS-filtered approach enhances the accuracy of predicting protein structures and identifying fold classifications.
• The method offers a valuable tool for predicting structural homologies and inferring protein function when sequence homology is insufficient.