CAMBRIDGE, Mass. & BERLIN—In June, Centogene published a white paper announcing its progress on an in-house artificial intelligence (AI) program designed to accelerate diagnosis in patients with rare diseases. The paper demonstrates how the company’s variant prioritization tool has not only accelerated the diagnostics process, but also outperforms other tools with regard to sensitivity and specificity for flagging “pathogenic” and “likely pathogenic” variants.
In the diagnosis of rare disease patients, variant prioritization is a vital step in discovering the causal variants used to identify disease-causing mutations. Variant prioritization accelerates and simplifies variant interpretation as the results enable the interpretation of variants of unknown significance. Prioritization scores enable the diagnosis of a patient and, indeed, rare disease diagnosis relies heavily on variant prioritization scores in order to determine which variants are likely to affect the function of genes.
This latest initiative, named CentoMD, is believed to be the world’s largest curated data repository. It includes epidemiologic, phenotypic and heterogenetic anonymized data of more than 420,000 patients and more than 7.3 million variants. Through CentoMD, Centogene is ideally positioned to leverage its “big data” for AI-driven accelerated diagnosis.
“Big data is the key enabler of artificial intelligence since AI systems need enormous data sets to train algorithms. The better and more comprehensive the data, the higher the predictive power and accuracy of results from artificial intelligence,” commented Dr. Volkmar Weckesser, Centogene’s chief information officer. “Centogene finds itself in this enviable position with our data repository CentoMD—our big data. We have demonstrated that by combining what we believe to be the world’s largest database of genetic information with an AI-based variant prioritization solution, we outperform other tools available and ultimately accelerate the diagnosis of rare disease patients.”
Carsten Ullrich, the company’s director of artificial intelligence, added: “Centogene’s diagnostic and pharmaceutical solutions rely on the extensive knowledge and insights held in our data repository. Artificial intelligence enables us to find relationships faster, draw more exact conclusions about relationships in the data and discover patterns that cannot be found with traditional methods.”
Centogene reports that since December 2018, the number of analyzed cases has grown by 16 percent to more than 360,000, and the number of total variants has increased by 26 percent to 9.3 million. The program now covers data from over 120 countries, underscoring the important ethnic diversity of Centogene’s knowledge base. By comparison, the freely available ClinVar contains 510,000 variants as of June 2019, while HGMD Pro contains 255,000 variants as of January 2019.
CentoMD’s data combines variant information with proteomic and metabolomics information, in particular for high-throughput genes and follows American College of Medical Genetics and Genomics guidelines for uniform variant classification. Through a highly qualified and standardized curation process, the program provides high-quality clinical interpretations of newly identified variants and ensures that changes in variant classification will be communicated and reflected in clinical interpretations in a timely manner.
As one of the largest rare disease companies worldwide, Centogene is focused on transforming clinical, genetic and biochemical data into medical solutions for patients. Its goal is to bring rationality to treatment decisions and accelerate the development of new orphan drugs by using its knowledge of the global rare disease market, including epidemiological and clinical heterogeneity and innovative biomarkers.
“Our detailed genetic, proteomic and metabolic analysis is the key to fueling the knowledge base of rare disease patient populations, helping to drive Centogene’s biomarker development program, and support our pharmaceutical partners in accelerating the development of orphan drugs,” explained Centogene’s CEO, Dr. Arndt Rolfs. “Ultimately our knowledge and expertise are for the benefit of our rare disease patients to help end the diagnostic odyssey that so many of them face. By following a strict data curation process, we are providing highly accurate data relevant to clinical diagnosis and decision-making.”