Specificity of Dendris


Dendris DNA microarray technology is based on a syndromic approach. Several pathogens are searched for in a single analysis, based on symptoms, allowing for improved diagnosis and management of patients.

A multiplex consensus PCR will amplify all pathogens present in the sample to be analyzed.

Hybridization will screen and identify pathogens present thanks to a combination of specific probes to each target.


The key points of DNA microarray technology are the sensitivity and specificity of the interaction between the probes and the target, which result from the design of the probes and the chemistry that binds the probes to the support. Our technology is based on the functionalization of the support with phosphorus dendrimers, which significantly increases the sensitivity of our chip. Dendris holds the exclusive license of this technology patented and developed by INSA and CNRS. Two founding members of Dendris are the inventors.


Dendris uses a machine learning algorithm to deliver the results of analyzed clinical samples. From a learning base of labeled samples (supervised) our algorithm will build its model. From this model it will be able to provide predictions on new samples in an automatic, reliable and fast.

The algorithm selected by Dendris is based on the combination of decision trees. This algorithm called Random Forest was proposed in 2001 by Leo Breiman, member of the Academy of Sciences in the USA. This program is widely used in artificial intelligence (AI) to solve prediction problems, especially in the field of health (1832 citations on PubMed in 2022).

Since the creation of Dendris in 2009, we have adapted and parameterized the Random Forest in order to better exploit the data generated by the DendrisCHIP biochip.

This constantly improved AI allows us to meet our customers’ expectations with optimal sensitivity and specificity of pathogen detection, making it a powerful tool for treating patients.

Samples transferred to the application enrich the learning base allowing the model to become more and more efficient and to enlarge the panels with new targets.

Our application and data are hosted on servers (dual site) certified as health data hosts (HDS) by our expert partner XEFI.

The HDS certification guarantees security through data encryption and a strict confidentiality policy.