BioStar: a data warehouse schema for integrating clinical and genomic data from HIV patients

In the field of biomedicine, it is becoming increasingly apparent that a huge amount of genomic and clinical data are being generated everyday. Although these data, stored in different formats at multiple sources, help us acquire more information about the patients, the scale and complexity of these datasets result in the challenges of integration of multiple biomedical datasets. Based on the traditional star schema and our previous Biostar schema we propose a hybrid schema called BioStar+ for HIV patient data modeling. By maintaining the original data space structure instead of disassembling it, this hybrid schema not only retains the original advantages from BioStar, but also improves the query efficiency. In addition, we describe the HIV patients data warehouse that we have developed based on the BioStar+ schema and discuss an analysis case based on this schema.