ATEM leads the way in lipid nanoparticle (LNP) characterization by transitioning from traditional, manual analysis to a groundbreaking quantitative approach. Utilizing AI-powered Cryo-EM, ATEM provides precise insights into LNP size, shape, and distribution—key factors that influence drug delivery efficacy and stability.
Our AI-driven method streamlines the entire analysis process, making it faster, more accurate, and cost-effective. This innovation not only supports large-scale stability studies but also offers our clients significant economic advantages by optimizing resource use and reducing time to market.
Understanding the intricate relationship between LNP morphology and its functional efficacy is central to our analysis. By detailing how structural attributes like size and shape directly impact the effectiveness of drug delivery, ATEM provides essential insights for optimizing nanoparticle design and formulation, aligning with the principles of structure to function in nanoparticle technology.
ATEM’s AI LNP solution is in full compliance with the 2022 FDA recommendations and guidance for industry regarding Drug Products, Including Biological Products, that Contain Nanomaterials. Our method ensures comprehensive analysis, covering size distribution, shape, and morphological characterization. It is capable of handling high cryoprotectant concentrations and is suitable for analyzing a wide array of substances, including mRNA and siRNA, in their undiluted, native states.
ATEM’s AI-enhanced Cryo-EM characterization method sets a new standard in the field, combining scientific accuracy with economic efficiency. Our approach accelerates the development of nanoparticle-based medicines, pushing the boundaries of what’s possible in drug delivery technology.
At ATEM we have developed a machine learning based algorithm to transform cryo-EM LNP images into quantitative morphological insights. For precise, reproducible and accurate characterization of thousands of LNP particles.
The FDA in its most recent industry guidance for “Drug Products, Including Biological Products, that Contain Nanomaterials – Guidance for Industry” (April 2022) recommends using cryo-EM for the characterization of multiple CQAs (Critical Quality Attributes).
ATEM now provides these readouts in a single assay at highest statistical robustness in a label free process.
Gain valuable insights to decipher the in-depth characteristics of your most promising formulations or obtain previously inaccessible perspectives on challenges.
Better understand bottlenecks in your scale-up or downstream processes and their effect on LNP drug product. Gain the ability to more effectively address and mitigate them.
Ensure the most advanced understanding into batch-to-batch consistency to ensure your processes and products always control for highest LNP drug product integrity.
Extend your IND filing with quantitative LNP cryo-EM data to showcase product quality, production consistency and in-depth understanding of its characteristics.
LNP Size & Size Distribution Characterization
ATEM’s machine learning algorithm can characterize the size and the size distribution of the LNP at single particle precision.
LNP Morphology Classification
The machine learning model classifies the particles into the main observed particle morphologies of Solid Core, Biphasics Split and Biphasic Dense (Blebbed) particles.
LNP Aspect Ratio
The Aspect Ratio (Shape Ratio) identifies the shape of the LNP, relative to a perfect sphere. Cryo-EM is the only method to provide this level of insights.
Based on a random draw study we can see that only 2.000+ particles the statistical significance becomes robust. This is important when drawing conclusions about less frequently present morphological classes.
< 1.0 nm
95% Confidence Interval
(Avg. Particle Diameter, 5000 particles)
Based on technical replicates of the same sample have been analyzed. And the result shows a Standard Deviation of sub 1 nm across all replicates. Additional information can be obtained from the application note.
< 1.0 nm
Standard Deviation
(Avg. Particle Diameter, 5000 particles)
Based on benchmarking the results of a trained human operator against the results generated by the machine learning model show human like results (>96% accuracy). Technical replicates demonstrate a 1-2% standard deviation on classification accuracy.
96.3%
Ø Classifiaction Accuracy
(Avg. Particle Diameter, 5000 particles)
Price per sample
Price per analyzed image data set for up to 500 images
Option 2 is aims at enabling partners with own cryo-EM hardware or established collaborations to benefit from transforming their images into quantitative insights. ATEM also offers the option to sell its specialized grids that enable up to 80% data collection success on the first grid.
We prepare everything else. We receive your sample, then optimise and test it before returning the data in a personalised way to meet your needs.