Data-mining interface. The "Filter database" data-mining interface allows users to select a gene set that meets specific expression, keyword, and sequence criteria. Input fields include a) differential expression quantified from stage- and tissue-specific expression microarray analyses, b) keywords included in functional annotation gathered by Ano-Xcel  from the ENSEMBL, NCBI non-redundant, PFAM, GO, and SMART databases, and c) presence of transcription factor binding sites and other conserved DNA sequences contained within promoter, 3' UTR, or coding regions of the An. gambiae genome. Each filter is imposed on the current gene set being examined, beginning with the entire An. gambiae genome, thus selecting and reducing the gene set in a stepwise fashion as genes matching previous filter criteria are eliminated by subsequent filters. The parameters specified here are those that are used in the prophenoloxidase case study described in the text.