{{selectedSearchMode}}{{selectedSearchMode=='Keyword Search'? ' / Upload Gene List' : ''}}

List of keywords:
Search Scope:
use exact match

Sample group:
more options... hide options

Filter out genes with counts < per million in at least samples


For a selected gene within a dataset, you can find other genes with correlated or anti-correlated expression (pearson correlation is used). This function is available from the gene expression profile page. Expression profile page is usually accessed from the gene set page, which shows the result of a search, such as keyword or differential expression search.

Search the full list of genes in the system for terms of interest. Use commas or spaces or new lines to search for multiple matches Eg. myb, suz12, ENSMUSG00000005672 will search for myb [or] suz12 [or] ENSMUSG00000005672.

You can also upload a list of genes here by entering a list of gene symbols or ids. Use search scope to ensure matches are made only within selected fields.

You can perform differential expression analysis here. Select dataset of interest, then sample group, followed by sample group items. Haemosphere will use R package limma to perform the analysis for microarrays and voom from limma and the package edgeR to perform the analysis for RNA-seq data. You can also view the R script and download the R objects used here.

Find genes with high expression in the selected sample group. Each gene is scored for the difference between the selected sample group and the highest value of the rest, and only genes with positive scores are returned. Also, genes with variance < 1 (on log2 scale) are filtered out, as well as those with max < (min of dataset +1). Note that for microarray datasets, the calculations are done on probes, then collated back to the gene level.

Plot gene vs gene. Selecting a dataset here will take you to the plot page with two arbitrarily chosen genes, where you can select genes from the dataset.

R function for differential expression analysis

The following is the R script used by haemosphere to run differential expression analysis. Note that this is read-only, and is provided for advanced users who would like to understand the underlying method used or to customise the analysis within R themselves (use the download feature under 'Datasets' to obtain the data). There is also an option to download all relevant R objects as an R binary file which can be loaded into R session using load() function. Just select the checkbox here first before running the analysis.

download R objects (your download will start after gene set page loads)