CHICAGO – In the last few months, Genedata has added support for digital pathology to its Profiler multi-omics translational research platform, and, through a partnership with a disease-specific organization, entered the nonprofit arena.
Genedata CEO Othmar Pfannes called the latter move, a data science project with the Crohn's & Colitis Foundation, part of the natural evolution for the 20-year-old, Basel, Switzerland-based company.
Profiler launched in 2015 as an enterprise platform for precision medicine that integrates next-generation sequencing and other omics data with clinical information to support patient and compound profiling processes for the biopharmaceutical industry.
Pfannes said that the Crohn's & Colitis Foundation is similar to a pharma firm in that the organization is conducting and funding R&D. The New York-based foundation has been collaborating with the software company for about a year to integrate and analyze multi-omics and clinical data from large groups of Crohn's patients in search of biomarkers that can indicate elevated risk of developing two particular complications that would require surgery: fibrosis and fistulizing Crohn's disease.
"If we can identify those complications early at diagnosis, then there would be the possibility to intervene early and to change the outcomes," said Andres Hurtado-Lorenzo, senior director of translational research at the Crohn's & Colitis Foundation.
They also are collaborating to apply machine learning to the foundation's data generated from this cohort to validate the findings of a long-term research study of pediatric patients with inflammatory bowel disease. The study, which has been going on at 28 US sites for the last 13 years, has led to the identification of gene signatures that can predict a Crohn's diagnosis.
This is the first time the Crohn's & Colitis Foundation has had machine learning for data analysis, according to Hurtado-Lorenzo.
"They collect large datasets — really big datasets — and ultimately they needed somebody to analyze this data," Pfannes said of the foundation. "We not only provided the software, but in this case, we also provided data analysis consulting."
Genedata is building up its in-house group of bioinformatics professionals, which currently stands at about a dozen.
"We have had data scientists for a long time, historically just supporting the [software] development," Pfannes said. "We now see a business opportunity [to offer services to] smaller biotechs."
In the scope of this partnership, the Crohn's & Colitis Foundation is functioning as a small biotech.
"They have lots of data. They need it to be analyzed and they don't have the resources from a personnel perspective," Pfannes explained. "We said we would provide software plus the human resources to do the data analysis."
Going forward, Hurtado-Lorenzo said that the Crohn's & Colitis Foundation will attempt to create tests based on its discoveries that might be able to prognosticate disease course. He said that the organization has had discussions with diagnostic companies on this front, though he did not name any of those potential partners.
For this test, the foundation and Genedata want to identify signatures to predict response to widely used biologics, particularly infliximab (Remicade), early in the disease course.
Hurtado-Lorenzo said that he was drawn to Genedata for its 20 years of experience working with pharma and biotech companies, though Crohn's & Colitis is new ground in that the organization is a nonprofit. This, Hurtado-Lorenzo surmised, could open up new possibilities for Genedata with other disease-specific foundations that are sitting on very large datasets.
"We have integrated microbiome data together with gene expression data together with proteomics and clinical data to create our prognosis model," Hurtado-Lorenzo said. He noted that Genedata applies not only machine learning but more traditional statistical analysis.
Already, the partners have moved on to proteomics to identify additional biomarkers for complications. That work so far has been based primarily on the analysis of intestinal biopsies using gene expression and transcriptome data. "We would like to [have] better understanding of the value of protein signatures in the blood," Hurtado-Lorenzo said.
The Crohn's & Colitis Foundation is looking to patent the method for analyzing intestinal biopsies using gene expression data and transcriptomics and also is preparing a scientific paper. Hurtado-Lorenzo declined to give specifics on the patent application until the organization submits it in the next month or two.
The collaboration with the foundation comes a few months after Genedata said that it had reached a major milestone in a multiyear software development effort with Merck, adding imaging capabilities to Profiler. Those firms developed the new features to meet the pharma giant's requirements for large-scale data management and analysis to support molecular profiling at multiple locations.
The software update features a new workflow for digital pathology that integrates imaging into multi-omic data, including next-generation sequencing, proteomics, metabolomics, lipidomics, and other types of omics data already in Profiler, as well as clinical and biomarker data to support more advanced analytics. Additional new workflows include neoepitope prediction and tumor mutational burden analysis for immuno-oncology, according to Genedata.
Initially, Profiler is supporting only pathology imaging, but the company plans on adding additional imaging modalities in the future, Pfannes said.
"Working closely with Genedata, we have generated a digital biomarker research platform that will enable Merck scientists to better utilize clinical and translational data to generate innovative ideas for new biomarkers and drug targets," Joern Peter Halle, Merck's immuno-oncology and external innovation chief, said in a statement back in October. He said that it would serve the company's goal of developing new precision therapies.
Merck would not comment specifically for this story, but Pfannes told GenomeWeb that the firms intend to continue development of Profiler as new AI techniques come online.
Merck has been a longtime customer of Genedata. Nearly four years ago, the pharma giant asked the tech vendor to make some modifications to meet its particular needs in designing better clinical studies, reflecting the industry movement away from blockbuster drug development to the search for precision therapies.
"You need to invest much more into finding the right patients than you used to do. We clearly support this type of concept," Pfannes said.
But the rise of precision medicine has meant the introduction of patient information earlier in the drug development process, which calls for heightened privacy and security protections.
"Historically, the preclinical space didn't worry too much about data security and data privacy issues," Pfannes said. "Historically, there was a very clear separation between clinical data and R&D in the preclinical space." That has changed.
In Profiler, Genedata tries to manage the "chain of custody," or the control of patient-specific information, the CEO said. Since the European Union's General Data Protection Regulation (GDPR) took effect in May 2018, any entity that handles electronic personal data of any resident of an EU country must be able to trace the flow of that information and destroy records should the individual withdraw consent for use. As a supplier of technology to a heavily regulated industry like pharma, Genedata likes to play up its security and privacy features, including its support for GDPR compliance.
Pfannes also noted that Genedata customers, particularly contract research organizations, are increasingly using Profiler as a collaboration platform, because researchers in disparate locations can share and integrate data and communicate their progress.