|Input options Explaining output|
Input optionsESCDb can be queried by "Gene based view" inputting a set of gene names or selecting a GO category (Fig. 1). Second option is using the databases via "Gene filtering view" where user can set filters for different datasets e.g. selecting genes bound by Nanog in three different experiments or combining it with expression datasets by selecting genes that go up in Nanog knockdown experiment (Fig. 2).
Figure 1 - Gene based view inputs Input can be given by list of genes to the Query field (A) or by using Go term Query (B) by typing part of a Gene Ontology category description (C). Selecting the targeted GO term from the appearing menu (D) and selecting the GO term Id (E) will make a query from the ESCDb with all the genes having selected GO annotation.
Figure 2 - Gene filtering view Dataset filters can be set to transcription factor binding datasets stating if the factor must be bound (B) or not bound (NB) (A) or to expression datasets stating if the gene must show significant (1.5 fold) up- or downregulation (Up/Down) (B). User can also select Do functional characterisation of the resulting gene set (C) option for finding statistically significantly overrepresented GO terms, KEGG and Reactome pathways and regulatory motifs for the list of genes. The profiling is done using g:Profiler tool.
Explaining outputOutput of the query is a table divided into two: transcription factor binding experiments and gene expression experiments. Both divisions are further colorcoded for mouse and human experiments.
Ticks on the left side of the table represent binding events and coloured boxes on the right hand represent expression values.
Pointing mouse to the boxes gives further information about the given gene in the specific dataset (img 3).
Figure 3 - Reading the output Hoovering over a transcription factor binding data box (a) additional information about the binding event is given (e.g. score of the peak). By pointing the mouse over expression dataset box (b) for each probe (if there are more than one) ratio, p-value and used probe ID are given. Each column header contains information about the dataset - cell line, PubMed ID (c)
When using Gene filtering view user can also select automatic gene set characterisation. These results as well as the genes belonging to defined gene group are listed in the bottom left corner of the output table (Fig. 5).
Figure 4 - Color coding of expression levels
Each table row explains one gene's expression in various datasets. Each box in the row can be further divided into 2 or more when the genes has more than one probe on that specific platform (e.g. column 3 in Fig. 4).
Box colors are following:
Figure 5 - Significant GO categories and genes