SAXS Assistant: Automated SAXS analysis for structural discovery in biologics and polymeric nanoparticles

Area of Science:

• Structural biology
• Biophysics
• Materials science

Background:

• Small-angle x-ray scattering (SAXS) is crucial for macromolecular structure analysis.
• Current SAXS data interpretation is time-consuming and subjective, limiting high-throughput applications.

Purpose of the Study:

• To develop SAXS Assistant, a Python script for automated SAXS data analysis.
• To streamline the extraction of structural parameters and machine learning-ready features from SAXS data.

Main Methods:

• SAXS Assistant integrates with BioXTAS RAW for data analysis.
• A multilayer perceptron regressor was trained on experimental data for Dmax estimation.
• Gaussian mixture model clustering was used for structural profile classification.

Main Results:

• The script validates data reliability using Guinier/PDDF R_g agreement.
• The Dmax estimation model achieved R² = 0.90 and MAE = 11.7 Å.
• Clustering provides quantitative shape-descriptive features and similarity assessments.

Conclusions:

• SAXS Assistant accelerates SAXS data analysis with quality control and ML-ready outputs.
• The tool enhances high-throughput analysis and aids researchers in biological and polymer fields.
• It provides flags for low-confidence results, ensuring reliable structural insights.