MDBiomarkers: A queryable biomarkers database integrating multiple serum and tissue datasets for Duchenne muscular dystrophy

BackgroundFit-for-purpose biomarkers are urgently needed in Duchenne muscular dystrophy (DMD). However, biomarker efforts in DMD have traditionally been hampered by a lack of reproducibility due to small sample sizes, confounders such as treatment and age, and discordant findings from different technologies. Moreover, there is no central resource to get an overview of cumulative published evidence. Hence, many researchers often start with new discovery studies, which are time-consuming and costly.ObjectiveBuild a dynamic, searchable, and easy-to-use biomarker platform for DMD.MethodsThousands of molecular (serum proteins and muscle mRNA) markers from multiple studies (28 analyses) were compiled. Findings were obtained from supplemental material of published manuscripts or by following standardized pipelines on available raw data. These findings were annotated with important attributes (e.g., age range, treatment, etc.). Evidence was aggregated around each biomarker's association with DMD, treatment, age, clinical outcomes, as well as other markers.ResultsThe interactive Shiny application on https://www.mdbiomarkers.com provides exportable summaries of serum protein and muscle tissue mRNA findings. This also permits new knowledge to be generated for nuanced meta-analyses, rather than being restricted by a single study's finding and p-value. A tutorial is provided on the website. This resource is planned to be continually updated with new/additional findings to fulfill the aim of a living biomarker resource.ConclusionsThe resource developed will reduce preparatory time to distill evidence around important biomarker candidates providing summary estimates around individual studies' effect sizes, help assess cumulative evidence, and help with experimental design of future experiments.