We used several interaction/pathway tools to identify genes and proteins interacting with the identified GWAS-relevant genes: STRING (Search Tool for the Retrieval of Interacting Genes/Proteins) [25], MINT (Molecular INTeraction Database) [26], KEGG (Kyoto Encyclopedia of Genes and Genomes) [27] and Pharm-GKB (The Pharmacogenomics Knowledgebase) [28]. For STRING analysis, we excluded text-mining from the active prediction methods and considered outcomes with the highest confidence (STRING score >0.900). For the MINT analysis, we considered Homo sapiens, as the reference organism, and interactions with a score greater than 0.50. For KEGG and PharmGKB, we considered interacting genes/proteins those involved in direct interactions: either involvement in the same catalytic reactions or immediately preceding/subsequent catalytic reactions with proteins encoded by GWAS-relevant genes. With these criteria, we identified 82 interacting proteins. Supplementary Figures 1, 2 & 3 portray the interacting networks based on study of AD, ND and OD, respectively.
