CHICAGO – With its introduction of a bioinformatics product called Biomedical Knowledge Base late last month, Qiagen is targeting a new demographic in the biomedical research space.
Released through the Qiagen Digital Insights (QDI) bioinformatics division, Qiagen Biomedical Knowledge Base is a collection of manually curated data that lists structured biomedical relationships between elements including genes, diseases, functions, targets, drugs, and chemicals to support graph analysis, artificial intelligence, machine learning, and other drug-discovery applications. It is the firm's first product specifically designed for data scientists.
Biomedical Knowledge Base can draw on Qiagen's curated database created 20 years ago to support the firm's flagship Ingenuity Pathway Analysis (IPA) product for interpretation of multiomics data, according to Marina Bessarabova, Qiagen's San Diego-based director of project management discovery.
"Now our strategy is to provide customers with direct access to this data so they can mine it their own way, not [necessarily] through the predefined way of Ingenuity Pathway Analysis," Bessarabova said. "It's opening up direct access to our content for data scientists."
Biomedical Knowledge Base supports analytics-driven drug discovery by allowing customers to combine the Qiagen database with their own analytics pipelines to feed third-party or custom algorithms for generating hypotheses around drug targets, drug candidates, and new indications for existing compounds. The product also can integrate Qiagen data with information from outside sources.
Qiagen adds a semantic layer on top of the data it aggregates.
Data scientists can use this technology to establish an analytical pipeline for predicting drug targets as well as to create interactive target dossiers that resemble dashboards or portals, according to Bessarabova.
She said that one early adopter is using the knowledgebase to build its own portal that integrates the Qiagen data store with external data sources, then analyzes everything together.
A key feature of Biomedical Knowledge Base is graph analysis to visualize relationships between numerous biomedical objects in the core Qiagen database. In an online demonstration to GenomeWeb, Bessarabova showed how Qiagen technology matches drugs to diseases and diseases to targets using EGF and EGFR in non-small-cell lung carcinoma as an example.
"We have information not only about the fact that there is a relationship, but also we know the causality," Bessarabova said. "And we know a lot of additional information about those relationships, for example, in what organism it was observed, what was the experiment setting, what was the tissue or organ where it was observed?"
The integration of this kind of information from multiple sources allows data scientists to discover insights that can lead to new hypotheses, she explained.
"Normally, this is disintegrated information," Bessarabova said. "We integrate it, we put it together with structure manually, which guarantees high quality, and then with all of this integrated information, customers can run their advanced analytic algorithms to generate hypotheses that otherwise would be hidden."
On its blog prior to the formal release of the Biomedical Knowledge Base product, the company described how it used the technology to build an algorithm to reconstruct pathways of COVID-19 from viral proteins in SARS-CoV-2. In this test, Qiagen was able to identify about 450 existing drugs that could potentially interfere with virus-host interactions. Some 54 were already involved in COVID-19 clinical trials, the company said.
Qiagen has not disclosed the names of its beta testers or early customers, since all are in the rather secretive fields of pharmaceuticals and biotechnology.
"We are looking to organize a user group [with] meetings and webinars moving forward where we would be inviting some of our customers to present," Bessarabova said. "At that point, we will give more information."
The firm did say that Qiagen Biomedical Knowledge Base will initially be available only in North America, Europe, Japan, and Australia. Bessarabova added that the company is not specifically targeting academic researchers or contract research organizations with Biomedical Knowledge Base.
According to Bessarabova, this is just the first of several planned Qiagen products aimed at data scientists, likely based on existing QDI datasets.
"The objective is to have an overall integrated solution that would provide direct access to all our content within QDI to data scientists and bioinformaticians," Bessarabova said.