| Searching by gene/genomic coordinates |
To retrieve transcription factor binding profiles for specific gene(s) or genomic region(s), a list of gene symbols (i.e. HHEX, ETV6, etc.) can be entered one per line. Genes can also be specified using wild cards, for example "ETV*" would retrieve all genes beginning with "ETV" (i.e. ETV1, ETV2, etc.). The field also accepts searches by genomic coordinates in BED format (i.e. chr start end). If genomic coordinates are entered, the checkbox "search by genomic coordinates" must be selected. Finally, a gene list of BED file can be directly imported for query using the "Import" option. The figure below illustrates a search using a list of 5 genes.
|
| Searching by transcription factor binding |
BloodChIP also allows users to retrieve all genes either bound by a combination of selected TFs or by any of the selected TFs. To retrieve genes that are regulated by multiple TFs across multiple cell types, the "AND" condition can be selected between each of the selected TFs. To retrieve all genes bound by any of the selected TFs, the "OR" condition can be selected. The example figure below illustrates how to retrieve all genes combinatorially bound by ERG, FLI1, SCL/TAL1, RUNX1, GATA2, LMO2 and LYL1 in CD34+ cells. If one is interested in retrieving all genes bound by any of the same set of TFs, the "OR" condition will be selected in each case.
The list of a genes can be further refined by using a fold-change filter between any of the available cell types. The same figure below illustrates how to retrieve only genes that are 2 fold higher in CD34+ compared with Megakaryocytes.
Note that the fold change used in this field is the raw fold change (i.e. A fold change of 2 will be a log2 value change of 1 as shown in the expression box plots in the table).
|
| The gene/TF binding table |
The genes retrieved by a query are displayed in a table underneath the results. Each gene entry can be expanded to show TF peaks associated with the gene. In each TF peak row, the binding status of all TFs in each cell type is shown. In the current configuration, the three cell types, CD34+ HSPC ("C"), megakaryocytes ("M") and SKNO-1 ("S") are shown for each of the seven TFs. A red box indicates that the particular TF in the particular cell type is binding at this peak. A blue box indicates that there is no binding and a white box indicates that the TF binding profile is unavailable for the particular cell type. Histone and DNase I hypersensitivity (DHS) profiles are displayed in the same manner. Specifically the number of digital genomic footprints (DGF) is also shown where a DHS site is present at the peak.
Note that when searching by TF binding, only peaks that match the TF binding condition specified in the search are shown, where as the default behaviour when searching by gene is for all peaks associated with the gene to be shown.
Apart from displaying TF binding information, the table also provide links to UCSC tracks for the visualisation of TF binding profiles at a specific gene or peak locus. Furthermore, for each gene, a link is provided to Expression Atlas. Clicking on each gene will also display the gene expression profile for the gene across different cell types and CD34+ sub-fractions which are shown on the right of the table.
|
| Export functions |
The BloodChIP interface also profiles a number of export functions, located at the top of the gene table. The export options are as follows:
(1) Export TF binding status for all genes in the table.
(2) Export gene expression profiles for all genes in the table.
(3) Export file for network visualisation in Cytoscape.
(4) Link to STRING database.
|
| REST interface |
|
The BloodChIP also provides a REST interface for programmatic access to the BloodChIP database. Instructions to use this interface can be downloaded here: BloodChIP_REST_tutorial.pdf
|
| How to cite |
| Chacon D, Beck D, Perera D, Wong JWH and Pimanda JE. BloodChIP: An Atlas of Genome-wide Transcription Factor Binding Profiles in Human Haematopoietic Stem/Progenitor Cells. Nucleic Acids Res. 2014 Jan;42(Database issue):D172-7. doi: 10.1093/nar/gkt1036 |
|
|
