New DMDD data released on Expression Atlas today reveals the effect of single gene knockouts on the expression of all other genes in the mouse genome. The gene expression profiles of 11 knockout lines have been derived from whole embryos harvested at E9.5, and the results can be compared with wild-type controls using an interactive online tool. Users can investigate which genes are differentially expressed as a result of a gene knockout, with the potential to uncover genes with similar roles or compensatory effects when a related gene is knocked out.

Data for additional lines will be released throughout 2017. The ultimate goal is to bring these molecular phenotypes together with the morphological phenotypes that have already been derived by the DMDD programme, to offer new insights about the effects of gene knockout on embryo development.


The knockout of Ssr2 in the mouse was found to affect the expression level of 325 genes in total, and this is one of the 11 new datasets that can be explored in Expression Atlas.

The differential expression of each gene is described using the log2 fold change – a measure that describes the ratio of gene expression in the knockout to the level of gene expression in a wild-type control. A negative fold change (shown in blue in the image below) means that the gene was expressed at a lower level in the mutant. A positive fold change (shown in red in the image below) means that the gene was expressed at a higher level in the mutant.


A visualisation of the level of differential expression of 8 genes affected by the knockout of Ssr2.
Eight genes that are differentially expressed due to a knockout of the gene Ssr2 (above a cut off log2 fold change of 0.4). Six genes are expressed at a higher level, while Mfap2 and Ssr2 are expressed at a lower level.


The interactive tool in Expression Atlas allows different cut-offs to be applied to the fold change, so the genes displayed can be restricted to those with a large differential expression. The image above shows the 8 genes with a fold change greater than 0.4 as a result of knocking out the gene Ssr2.

The tool can also be used to visualise the data in graphical form. The plot below shows the fold change for each gene, allowing the user to quickly ascertain the extent to which a gene knockout caused differential expression of other genes. All 325 genes considered to have a significant change in the level of gene expression are plotted in red, with the rest shown in grey.


Graphical visualisation of the fold change for each gene in the mouse genome, following knockout of the Ssr2 gene.
A graphical visualisation of the fold change for each gene. The outlier with a fold change of -3.5 is the gene Ssr2, which has a much-reduced expression level in an Ssr2 knockout embryo.


The full list of lines with data currently available is: 1700007K13Rik, 4933434E20Rik, Adamts3, Anks6, Camsap3, Cnot4, Cyp11a1, Mir96, Otud7b, Pdzk1 and Ssr2.

The full dataset for any line can be downloaded for further analysis, while the individual line pages on Expression Atlas integrate the DMDD data with other pre-existing data, in cases where a gene has already been shown to alter expression.


Knowing the ‘normal’ expression of genes during embryo development is key to understanding the differences that occur due to genetic mutations.

As part of work to understand the underlying transcriptional processes in developing embryos from knockout mouse lines, DMDD has now released a gene expression profile for wild-type mouse embryos between E8.5 and E10.5. The new dataset reveals the typical expression profile of genes during this crucial period of embryonic development, including their abundance, and when they are turned on and off.


RNA-seq has been used to establish the expression profile for whole, wild-type embryos at each somite number between 4 and 36 (excluding 29 – 33). This range corresponds roughly to the period E8.5 – E10.5, a vital period during which many organs and systems begin to develop.

The resulting data is now available in Expression Atlas. It’s a temporal baseline expression reference derived from wild-type embryos, which adds to EBI’s established resource to give a more complete picture of gene expression during embryonic development.


The wild-type baseline helps us to answer the question “what does ‘normal’ whole-embryo gene expression look like during development?” This is hugely important, as we can only really begin to explore what is abnormal once we know what is normal.

More specifically, the baseline highlights patterns in the way different genes are usually expressed as an embryo develops: when they are turned on and off; their abundance and whether their expression is covariant with other genes. Example expression profiles are shown below for Nacad and Pdzk1, indicating that at this depth of sequencing Nacad is switched on during somitogenesis and Pdzk1 is switched off.

Click to view larger image
Expression profiles of the Nacad and Pdzk1 genes with increasing somite number. The white boxes indicate no expression at a cut off of 0.6 fpkm (fragments per kilobase per million). The numbers in the boxes give the level of expression in fpkm, with bluer boxes indicating a higher level of expression.


For DMDD, the new dataset will underpin work on molecular phenotyping, by allowing us to understand whether the expression patterns of mutant embryos are significantly different from the wild-type. The ultimate goal is to allow users to correlate a given gene with the physical manifestations of its knockout in the developing embryo, and the underlying transcriptional processes.

The relationship between gene, morphological phenotype and molecular phenotype in the DMDD programme.
The DMDD database will ultimately allow correlation between genes, morphological phenotypes and molecular phenotypes (based on transcriptional processes).

However the data is a valuable resource for any researcher interested in gene expression during embryonic development, and is free to use. You can explore the data further in Expression Atlas.