How to use HeartEXpress

HeartEXpress is a web database for the study and visualization of integrated expression data related to heart development. Its main purpose it to help researchers in the (i) study of transcriptional responses in different experimental conditions; (ii) detection of co-expression and potential co-regulation and (iii) identification of potential functional relations between genes.

The HeartEXpress Home page (Fig.1) presents an overview of the features of HeartEXpress. In the top frame, links are provided to the other pages: Gene Expression Analysis, Integrated Datasets and Help.


Figure 1: HeartEXpress Intro page

HeartEXpress currently possesses two types of organisms (Homo sapiens and Mus musculus), with different types of clustered datasets which includes in vitro differentiation & reprogramming and in vivo. Links to these datasets are provided on Gene Expression Analysis (Fig. 2). These expression datasets contains genes that have up to 60% of their measurements missing. Human in vitro differentiation (Fig. 2A) contains 19736 genes and 33 experimental conditions; Human in vitro reprogramming (Fig. 2B) contains 19559 and 13 experimental conditions.


Figure 2: Gene Expression Analysis page - Human Datasets

For mouse datasets, the experiments were divided into three sub-categories: Mouse in vitro differentiation (Fig. 3A) contains 20881 genes and 18 experimental conditions; Mouse in vitro reprogramming (Fig.3B) contains 20657 genes and 34 experimental conditions; Mouse in vivo (Fig. 3C) contains 20904 genes and 29 experimental conditions.


Figure 3: Gene Expression Analysis page - Mouse Cluster Datasets

Integration of human and mouse data has been performed by using NCBI homologene file to identify the homolog genes between species. This integrated data set (Fig. 4) contains 16445 and 131 experimental conditons


Figure 4: Gene Expression Analysis page - Human & Mouse Cluster Datasets

Clicking on a dataset leads to the visualisation and exploration tool (shown in Fig. 5), which is a modified version of GeneXplorer application. Here, the user can examine differential expression of a gene or set of genes across different experimental conditions as well as the co-expression of genes. In the left frame, a miniature heat map of the cluster expression data is shown. Using the mouse pointer, a range of genes can be selected and their expression values will be enlarged in the right frame (Zoom). The expression data is visualized as matrix (or heat map) formed by rows representing genes, and columns representing the experimental conditions. On the right side of the matrix, the corresponding names and descriptions are displayed.

The expression changes are colour-coded by a blue to yellow colour gradient, corresponding to decreased (blue) or increased (yellow) expression compared to the reference sample or average expression across the measurements included in a experiment. A colour-bar on the right side indicates the range of fold changes. No differences in expression are represented by black squares. Missing expression values are represented by grey squares.

The user can hover the mouse pointer over the expression values, to obtain the gene description (Gene), experimental condition (Sample) and differential expression (Log2 fold change) of that gene. This information is displayed above the Zoom frame.


Figure 5: HeartEXpress Visualization and Exploration tool

The user can query for the expression values of a specific gene or a set of genes. To do so, the user has to input the desired gene Symbols (as example: Mef2c, Tbx5, Nanog and Pou5f1 - yellow elipse) or their specific description in the "Search for" box, define in which fields he wants to look for it and then press the GO button (Fig. 6). It is possible to download the queried results into a .txt file (blue arrow). Please note that the list of gene have to be separated by the enter key.

When the user performs the search, it will appear the queried genes and their expression values. Performing this search will enable the user to easely assess if the queried genes present similar or distinct expression patterns. To observe a correlation of one gene with other genes, the user as to select one gene (as example: Gata4 - yellow arrow).




Figure 6: Search by Gene Identifier.

Once the heat map with the expression data is displayed, the user can analyse what genes are correlated with the selected gene. Top 20 genes are displayed, presenting a high correlation coefficient with the selected gene, with a minimal coefficient correlation of 0.3 (Fig. 7). Hovering the mouse pointer over the orange bars adjacent to the genes identifiers displays the correlation coefficient as tooltip. Results can be download into a .txt file.




Figure 7: Set of genes with correlated expression changes

The other fields in the upper frame (Radar and Zoom) can be used to control the size of the two heat maps displayed on the webpage. Radar increase or decreases the miniature heat map in the left frame, while Zoom vary the size of the heat map displayed in the lower frame.


Contacts
For any query or suggestion please contact:

Rui Machado
Systems Biology and Bioinformatics Laboratory
Centre for Biomedical Research (CBMR)
University of Algarve
Faculdade de Ciências e Tecnologia
Campus de Gambelas
8500-139 Faro, Portugal