CAUSALdb was established using a JAVA-based web framework. We stored the partitioned summary statistics and fine-mapping results of causal blocks in flat files. The information pertaining to curated GWASs and potential causal variants was stored in MySQL for quick retrieval. The annotation information was indexed and stored using MySQL or Tabix (47). We generated highly interactive Manhattan and LocusZoom-like plots for users to inspect causal blocks using jQuery, D3.js, and related JavaScript modules. The overall architecture of CAUSALdb is shown in Figure 1.
