VAS is a Vironova proprietary tool for computer-assisted analysis of transmission electron microscopy (TEM) images in a GMP-certified environment. It is suitable in biotechnology and life science environments requiring procedures and controls that ensure integrity, confidentiality, authenticity, and reliability in their electronic records in accordance with FDA 21 CFR part 11.
VAS has machine learning capabilities, making it possible to set up methods for automatic nanoparticle classification. The machine learning method, Support Vector Machines (SVM), is a supervised learning method. This means that the classifier must be trained on a set of manually classified particles before it can be applied to new, unclassified particles. To do this discrimination, it uses a predefined combination of features, including shape, size, and intensity profiles.
The trained classifier has proven capability of discriminating between filled, empty, and intermediate particles. It can be applied to all particles in an image or to individual particles.
Significant time can be saved with VAS. This example shows the time required for manual analysis of liposome lamellarity in twenty images of an average quality sample. Using VAS instead of the manual imaging, detection, classification, and reporting can save your organization twelve hours per sample.
Virus particle packaging analysis with cryoTEM in combination with VAS is a patented method by Vironova and the world's first validated method for full/empty/intermediate AAV capsid ratio analysis. This method is recommended for batch release testing and quality control of AAV-based gene therapy products.
Representative cryoTEM images used for packaging analysis in VAS
Detected image (left) displaying the classified AAV particles overlaid with red, blue, and yellow circles. Principal component
analysis of each AAV particle’s radial density profile was performed using VAS and is displayed in the cluster plot (right).