AssignSLP_GUI, a software tool exploiting AI for NMR resonance assignment of sparsely labeled proteins

Area of Science:

• Biochemistry
• Structural Biology
• Computational Biology

Background:

• Uniform isotopic labeling (¹³C, ¹⁵N) is standard for protein resonance assignment but challenging for mammalian proteins with native glycosylation or disulfide bonds.
• Sparse labeling offers a cost-effective alternative for such proteins, but requires sophisticated computational methods for resonance assignment.

Purpose of the Study:

• To introduce AssignSLP_GUI, an updated software package for assigning NMR resonances from sparsely labeled proteins.
• To demonstrate the utility of AssignSLP_GUI by applying it to a sparsely labeled glycoprotein domain.

Main Methods:

• Utilizing a combination of NMR parameter measurements and structure-based predictions.
• Employing computational tools including neural networks for chemical shift prediction and genetic algorithms for assignment searches.
• Integrating AI-based structure prediction for enhanced assignment accuracy.

Main Results:

• AssignSLP_GUI successfully assigns resonances for sparsely labeled proteins.
• The software integrates advanced computational techniques for improved assignment efficiency and accuracy.
• Application to a CEACAM1 glycoprotein domain showcases the package's capabilities.

Conclusions:

• AssignSLP_GUI provides a valuable tool for structural determination of sparsely labeled proteins, particularly those from mammalian systems.
• The software enhances the accessibility of NMR-based structural studies for complex proteins.
• This work advances computational approaches in protein structure determination.