ACTRIMS 2026 | Applications of the MedCP framework: comorbidity mapping and identifying prodromal features of MS

The core advancement in our tool, as for other tools, is the attempt to solve an accessibility gap. So, meaning that sometimes between a researcher’s question and a reasonable answer or a data-driven answer, there is a barrier that is built upon the bricks of the technical skills that are necessary in order to analyze, to access, even not to analyze, just to access sometimes the real-world data or, in general, big data from biomedical sources. So, these systems actually accelerate and eliminate, in some cases, these problems. So, what we showed at ACTRIMS are just two examples of applications that can be, of studies that can be done with MedCP. One was comorbidity mapping. So, we asked the system to map the co-occurrence between autoimmune and neurodegenerative diseases in real-world data from the EHR. And then to match this epidemiological result with biological association, biological connectivity between these entities. And we show that there is indeed a correlation between the real-world data and biomedical a priori knowledge. All of this analyzing a 3 million patients’ data records and throwing the results out in almost five minutes. So, this is impressive per se. And another example that we showed at ACTRIMS was focused on MS, so we asked through MedCP to identify prodromal conditions in MS. So, we defined a cohort asking the system to select people that had at least five years of data before the first diagnosis of multiple sclerosis in our system, then to match one-to-one to controls on age, sex, race, and healthcare utilization. And then after selecting this cohort of 1,500 people and matched controls, we identified 146 conditions that were enriched in people with MS in the prodromal phase compared to the matched controls. And then to show an example of how using SPOKE, how using external knowledge could enrich the results in the same exit analysis, we focus on migraine and confirm that there is representation of migraine diagnosis in people that will in the future be diagnosed with MS. And then we show that there are biological connections that are driven by calcium signaling, neuroinflammation, ion channels that link at the biological level migraine with multiple sclerosis. So, these are just two examples, but many more can be done through our system. And yeah, hopefully, when it will get published, many other researchers will test it for their questions.

This transcript is AI-generated. While we strive for accuracy, please verify this copy with the video.