New approaches to aggregation analysis by CryoTEM

Direct visualization and objective quantification of particle aggregation

Particle aggregation is a critical attribute across biologics, viral vectors, and nanoparticle-based systems. Even low levels of aggregation may influence product stability, biological activity, immunogenicity risk, and batch consistency.

Conventional bulk analytical techniques provide global metrics but do not directly reveal the structural organization or morphology of aggregates at the single-particle level.

CryoTEM enables direct visualization of particle populations in near-native conditions, preserving structural integrity while allowing detailed population analysis. Beyond qualitative imaging, quantitative image analysis transforms CryoTEM into a powerful tool for objective aggregation assessment.

To address different analytical needs, QuTEM has developed two distinct and complementary approaches for aggregation analysis: a Classification Tool and a Statistical Tool

For more information, click here to check our webinar dedicated to this method: A new approach to AAV aggregation analysis

Method 1 - Classification Tool

Particle-level categorization

With the classification tool, the particles are categorized into defined populations based on a visual assessment. The classifications can directly be correlated to the images.

Automated identification of individual particles and clusters
Classification into monomers, dimers, and higher-order aggregates
Quantification of each aggregate category
Particle classification at an individual level

This approach provides detailed structural insight and precise categorization of the specimen's aggregation state, with a clear visualization of the individual particles, clusters and aggregations to get a clear overview of the different populations present within the sample.

Method 2 - Statistical tool

Population-level Analysis

The Statistical Tool provides an objective, population-level assessment of aggregation. The distance to the nearest neighbor is calculated for each particle in the image and compared to a fitted reference distribution generated for the same on-grid particle concentration. This approach corrects for random or preparation induced proximity, ensuring that observed clustering reflects true aggregation rather than artifacts.

Global evaluation of the entire particle population
Model-based comparison to a fitted reference curve
Detection of subtle shifts in aggregation behavior
Reduced user bias through statistical analysis

This approach enables robust, objective comparison of aggregation levels and is particularly valuable for trend analysis, stress testing, and batch-to-batch evaluation, making it particularly suited for in-depth characterization and comparison studies.