TOOLS & FEATURES
tools for complex queries
by biology researchers - developed by computer scientists
Graphical user interface.
GENEVESTIGATOR offers a rich graphical interface to query and interpret a very
large database of professionally curated expression data. Its
tools help you to find relevant conditions for your genes of
interest, to find genes having specific properties (e.g. biomarkers, targets),
or to identify gene expression modules that are co-regulated over
selected conditions. The tools let you analyze either individual experiments (Single Experiment Analysis)
or thousands of experiments simultaneously (Compendium-Wide Analysis).
SINGLE EXPERIMENT ANALYSIS tools
Samples tool. With this tool,
explore the expression level of genes across samples from any
experiment selected from the database. Compare the expression of
multiple genes and find detailed information about samples in which
expression is particularly high or low.
Powerful and extremely fast tool to run differential expression
analysis between two groups of samples from a chosen experiment. The
user can choose and experiment, define samples belonging to each
group, and run LIMMA with multiple testing correction
CONDITION SEARCH tools
Anatomy tool. Visualize the
expression of genes across many tissues and cell types. Example
shown: expression of the PAX2 gene across 246 different human tissue
types. The plot is sorted by highest to lowest expression. For each
boxplot, a detailed view is available to see expression in each
Cell Lines tool. Visualize
the expression of genes across a large collection of cell lines.
Example shown: expression of the PCSK4 gene across 416 different
human cell lines. The plot is sorted by highest to lowest expressing
cell line for this gene. The detailed view shows the sample level
expression of the first category.
Cancers tool. Visualize the
expression of genes across a large collection of different cancer
types and compare with normal tissues or with metastasic or xenograft
samples. Upper section: Boxplots of gene expression of the ERBB2 gene
across several subtypes of breast cancer. Lower section: Detailed
view of the selected boxplot. With the Neoplasms tool, you can
visualize the expression of genes of interest across more than 1,000
different subtypes of cancers and include over 250 different normal
tissues as controls into the same plot.
Perturbations tool. With this
tool, you can easily identify conditions that significantly affect
the expression of genes of interest. This screenshot displays the
response of three probe sets representing the human SOX4 gene to 835
different experimental conditions. . The results have been filtered
by p-value and fold-change to extract the most relevant conditions.
In the lower section, you can drill down to the sample level for each
condition described above and get for instance clinical parameters.
Development tool. Expression
of selected genes from mouse across different stages of development.
For each stage, expression values and standard deviations are
calculated from all microarrays annotated for that particular stage.
GENE SEARCH tools
Anatomy / Cell Lines /
Neoplasms tools. These three tools have an identical layout and
allow searching for genes specifically expressed in chosen tissues,
primary cells, cell lines or cancers. In this example, we searched
for genes specifically expressed in fallopian tube neoplasm by
screening against 1734 tumor type categories. Approximately 60 genes
were found to be highly specific for this cancer.
Perturbations tool. This is
an example of a biomarker search using the Perturbations tool from
the Gene Search toolset. Crohn's disease and ulcerative colitis are
chosen as target categories (for up- or downregulated genes), all
other conditions are chosen as base (unchanged). We obtain a list of
genes dysregulated in these two diseases but minimally changed in all
Development tool With this
tool, you can easily and rapidly identify genes expressed at a
particular stage of development of an organism.
RefGenes tool. Identification
of genes having the smallest expression variance across 26,075 human
samples (Affymetrix 133 Plus 2 arrays). The two boxplots in the upper
section represent, as a comparison, the expression distribution of
PPIA and B2M (two commonly used reference genes for RT-qPCR) across
the same set of samples.
SIMILARITY SEARCH tools
Hierarchical Clustering tool.
Hierarchical clustering of the expression of 32 genes throughout a
time-course experiment. Leaf ordering was applied to achieve a better
representation of the chronological events. Two main clusters are
Biclustering tool. The
biclustering tool helps you identify genes that are up- or
down-regulated over a subset of conditions (rather than over all
conditions). Here we show a bicluster analysis of 200 genes across
555 conditions. 377 different biclusters were identified. One of them
is shown here.
Co-Expression tool. This tool
allows the identification of the genes most correlated to a target
gene across any chosen set of experiments. The results are calculated
on-the-fly, making it very flexible and intuitive. This screenshot
shows the identification of the genes most correlated with the human
TRDN gene across 1487 different types of tissues and cancers. Genes
mutually correlated above a given threshold are linked, revealing
distinct clusters among the most correlated genes. Combining this
tool with the Perturbations tool above allows users to find
biologically very meaningful co-regulation clusters.
Signature tool. With this
tool, you can screen the complete database to find results that are
similar to those you have obtained in your own experiment. Simply
enter a vector of expression (gene identifiers and corresponding
expression values from qPCR, microarrays or RNA-seq) and run the
search. GENEVESTIGATOR searches for those conditions in which these
genes are regulated the same way.
Gene Set Enrichment tool.
Enter a gene set and easily find out to which pathways, biological
processes or diseases it is related. In the top table, a list of all
significant gene sets is displayed. In the bottom table, the genes
populating the gene sets selected are displayed. A Venn diagram
allows you to visualize up to four gene sets and to select genes that
overlap between these sets.
Selecting experiments of
interest. You can easily create a selection of data you would like
to work with. Choose either entire experiments (chosen by keyword or
by therapeutic area), or select all samples having a particular
description (tissue type, disease, cell line, cancer type, genotype,
etc.). The selection by keyword or therapeutic area is cross-platform
and cross-organism, thus allowing to find all experiments related to
a particular topic.
Filtering options. All
experiments in GENEVESTIGATOR have been carefully and precisely
annotated by our curators. Users can select experiments or samples at
a very detailed level, or can filter results and check the effect of
clinical parameters. The above screenshot shows how a selection of
samples can be filtered by parameters like age, sex, disease state,
Viewing experimental details.
Across all GENEVESTIGATOR tools, further details about each gene or
condition are available in mouse-over windows. In this example, the
exact experimental conditions for doxycycline study 2 are described,
including the genetic background and other parameters. Web links to
the original repository and to publications describing this
experiment are also provided.