A novel bioinformatics SaaS platform to identify and classify the pathogenicity of single genomic variants and oligogenic variant combinations for the diagnosis and treatment of genetic diseases

Projektinformationen

eVai

ID Finanzhilfevereinbarung: 190164416

DOI

10.3030/190164416

Projekt abgeschlossen

EK-Unterschriftsdatum 28 April 2022

Startdatum 1 April 2022

Enddatum 31 März 2024

Finanziert unter

• The European Innovation Council (EIC)

Gesamtkosten

€ 2 403 625,01

EU-Beitrag

€ 1 682 537,51

Koordiniert durch

ENGENOME SRL
Italy

Periodic Reporting for period 2 - eVai (A novel bioinformatics SaaS platform to identify and classify the pathogenicity of single genomic variants and oligogenic variant combinations for the diagnosis and treatment of genetic diseases)

Berichtszeitraum: 2023-04-01 bis 2024-03-31

Zusammenfassung vom Kontext und den Gesamtzielen des Projekts

Genetic diseases affect over 300 millions of people worldwide and their diagnoses increasingly depend on the analysis of the entire genome. However, despite the advances in reading DNA, the identification of the disease-causing variants is lagging behind, and diagnosis of genetic disorders is only achieved in less than half the cases on average.
Our innovative SaaS platform is the first and only monogenic and oligogenic variant interpreter that can drastically uplift the current diagnostic rates by facilitating the discovery of pathogenic variants still unknown. eVai combines AI and bioinformatics to assess the pathogenicity of single and multiple genomic variants with accuracy and speed: starting from the list of patient’s genotypes and phenotypes, this SaaS outputs the list of pathogenic variants ranked by probability of causation and the suggested diagnosis.
.

Arbeit, die ab Beginn des Projekts bis zum Ende des durch den Bericht erfassten Berichtszeitraums geleistet wurde, und die wichtigsten bis dahin erzielten Ergebnisse

Bioinformatic algorithms and AI-based models have been integrated in the finely engineered cloud-based architecture of the eVai SaaS to provide a comprehensive, robust and fast analysis of genomic data. EVai is the first and only monogenic, digenic and oligogenic variant interpreter available on the market. The software can now analyze row sequencing genomic data with validated state-of-the-art technology to obtain the list of different classes of genomic variants occurring in a patient (up to millions per single patient). Once this step is completed, eVai enriches the variants identified with information from dozens of “omic” resources and applies international guidelines to classify variants according to pathogenicity. Finally, through the layer of artificial intelligence we have developed, it can leverage clinical and familiar information, together with monogenic, digenic and oligogenic interactions to identify the genetic cause of the patients’ diseases.

As a final step, eVai pinpoints the suggested diagnosis for each patient, consisting of a single variant All the features integrated in eVai were finely validated and benchmarked on public datasets and on real datasets obtained within established collaborations with different partners (main hospitals and institution in the field).

To further support the genomic community, we have released two additional resources. First, OliVer (available at oliver.engenome.com) is a database of digenic and oligogenic combinations, with insights obtained through our ML model. Second, we have released VarChat (accessible at varchat.engenome.com PMID: 38579245), the first generative AI-based assistant to identify and summarize key literature about variants. Both these tools are freely available as web-based applications and are natively integrated within the eVai software.

Fortschritte, die über den aktuellen Stand der Technik hinausgehen und voraussichtliche potenzielle Auswirkungen (einschließlich der bis dato erzielten sozioökonomischen Auswirkungen und weiter gefassten gesellschaftlichen Auswirkungen des Projekts)

eVai software is intended to be used for diagnostic purposes. enGenome is therefore undergoing a certification process to be compliant with the new European Regulation for In Vitro Diagnostics medical devices (IVDR 2017/746) and is fulfilling all the requirements to succeed in this process (e.g.: ISO 13485:2016 compliance, Software clinical validation, Technical documentation drafting). Despite the lack of notified entities to designate new IVD devices, we’ve established our certification path and entered in the review process, that is expected to finish by the end of the year.

eVai is the first and only monogenic, digenic and oligogenic variant interpreter, can analyze raw sequencing data and apply a layer of artificial intelligence to identify the candidate Suggested Diagnosis by integrating clinical and genomic information.
During the project, we released two freely accessible resources to further support the genomic community: oliver (oliver.engenome.com) a database of curated digenic and oligogenic predictions, and VarChat, the first genomic assistant to identify and summarize key literature about variants. Both OliVer and VarChat are natively integrated within the eVai software.

To protect our innovation and its key distinctive features, our IP strategy has been focused on the protection of eVai’s key features (i.e.: the monogenic ML-based variant prioritization, the digenic ML-based variant prioritization and the digenic mechanism prediction). Our patent portfolio includes two patents granted in Italy and pending in Europe US and China, and a third patent request filed in Italy.

eVai success will significantly depend on our internationalization capability: despite the software can be deployed abroad and can scale its capabilities, correct timing and commercialization strategy will play a key role in our success. Penetration strategies in new markets require a deep knowledge of regulations and commercialization rules and the company should be ready to scale-up to face new needs.

eVai Highlights

eVai, a SaaS platform for the interpretation of single variant and variant combinations