Teiko makes cytometry batch normalization simple for large-scale global trials

We’re announcing the integration of advanced batch normalization into our core pipeline. This is a significant step forward in simplifying cytometry for global clinical trials.

Current normalization tools require users to fine-tune default parameters. Translational scientists are (rightly) worried that if they set the wrong parameters, the batch normalization won’t run as planned. And as a result, cytometry users have either skipped batch normalization or struggled mightily to get things working. That era is over. With our integrated solution, our scientists do the heavy lifting of fine-tuning parameters, to make sure your results reflect biological reality.

The enhanced batch normalization delivers:

• Speed and scale: We easily support large-scale 1,000+ specimen trials and enable rapid turnaround of trial-ready results.
• Precision assurance: In our internal testing, we’ve delivered the lowest inter-batch coefficient of variation (CV) to date, ensuring reproducible data critical for high-stakes clinical trials.
• Real-time trial batch normalization support: When you run real-time with Teiko, we can apply batch normalization to each sample as it rolls off the production line. For time-sensitive trials, this is a great way to immediately dive into your data while minimizing the batch effect of a real-time sample.

‍

Building on this, we are set to introduce intuitive dashboard features that will make batch normalization clear and accessible for drug developers. Translational scientists’ other major concern is about the “black box” nature of batch normalization.

Our planned features will make batch normalization easily accessible and interpretable (without getting a PhD in math):

Visualization of batch correction metrics: on app.teiko.bio, we will display batch correction metrics, such as inter-batch CV, in real time, allowing users to see exactly how data is standardized. Built-in guides and visual cues will explain each step of the normalization process, eliminating the need for specialized expertise and addressing hesitancy about algorithm outputs.