Improving Protein Structure Prediction Using Integrative Cryo-EM and Ion Mobility Mass Spectrometry Modeling

Area of Science:

• Structural Biology
• Computational Biology
• Biophysics

Background:

• Accurate protein structures are vital for understanding cellular functions and drug discovery.
• Existing methods like cryo-EM and mass spectrometry provide valuable but often incomplete structural data.
• Low-resolution cryo-EM maps and ion mobility mass spectrometry (IM-MS) collisional cross-section (CCS) values lack atomistic detail.

Purpose of the Study:

• To develop an integrative scoring function, CRIM (cryo-EM + IM-MS), to enhance monomeric protein structure prediction.
• To leverage complementary low-resolution cryo-EM density and IM-MS CCS data as restraints within the Rosetta framework.
• To improve the accuracy of atomic coordinate determination from sparse or low-resolution experimental data.

Main Methods:

• Developed CRIM, a Rosetta scoring function integrating Rosetta REF2015 energy with CCS agreement (PARCS) and electron density (elec_dens_fast) terms.
• Tested CRIM on simulated data for 60 monomeric proteins.
• Validated CRIM on an experimental dataset of 54 proteins using available cryo-EM maps or CCS values.

Main Results:

• CRIM improved protein structure prediction quality, reducing mean RMSD from 3.65 Å to 2.90 Å and increasing mean TM-score from 0.88 to 0.90 on simulated data.
• On experimental data, CRIM lowered mean RMSD from 6.65 Å to 4.38 Å and raised mean TM-score from 0.73 to 0.79.
• CRIM produced competitive predictions compared to AlphaFold3, outperforming it on challenging cases with sparse restraints.

Conclusions:

• CRIM effectively integrates low-resolution cryo-EM and IM-MS data to improve monomeric protein structure prediction.
• The method offers a practical approach for enhancing structural models when detailed experimental data is scarce.
• CRIM is available in the Rosetta software suite, facilitating its use in structural biology research.