Designing the database for microarray experiments metadata

摘要

Advancements in both computer science and biotechnology opened way for an unprecedented amount and variety of gene expression raw data to appear in the open access. It is sometimes worth to rearrange and unite data from several similar gene expression studies into new case-control groups to test new hypothesis using available data. Unfortunately, most popular gene expression databases, such as GEO and ArrayExpress, were not designed to allow these cross-study procedures. In order to locate comparable samples in different studies numerous steps are required including gathering additional sample metadata and their standardization. Specialized databases are developed by investigators in their own fields of interest to reuse the processed data and create different case-control groups and test multiple hypothesis. Here we present a detailed description of the specialized database creation along with its use case which is 35 gene expression cDNA microarray datasets containing expression data on more than 1000 biological samples on human placenta under conditions of preeclampsia. Samples contain sufficient metadata for them to be merged into relevant cross-experiment case-control groups for further integrative analysis.