Single-Cell Signature Explorer

Single-Cell Signature Explorer is a package of four softwares dedicated to high throughput signature exploration in single-cell RNAseq analysis. Single-Cell Signature Explorer  is fully compatible with Single-Cell  Virtual  Cytometer and Single-cell Multilayer Viewer .

 A video how to use this set of softwares is available as supplemental data in Nucleic Acid Research website.  How to install softwares and packages to use the Viewer & Combiner on windows is now available on youtube. Please note that this video covers the installation of R, Rstudio, Rtools, that are not supported by us.

Single-Cell Signature Explorer is directly compatible with Single-Cell Virtual Cytometer , Single-cell Multilayer Viewer and Single-Cell Spatial Explorer .

1. Single-Cell Signature Scorer computes for each cell a signature score.

2. Single-Cell Signature Merger collates the signature scores table with t-SNE/UMAP/other coordinates.

3. Single-Cell Signature Viewer displays signatures scores on a t-SNE/UMAP/other map.

4. Single-Cell Signature Combiner displays the combination of two signatures scores on a t-SNE/UMAP/other map.

If you enjoy using Single-Cell Signature Explorer and find it useful, then please cite Fred'Softwares address and "Single-Cell Signature Explorer for comprehensive visualization of single cell signatures across scRNA-seq datasets. Nucleic Acids Research, gkz601, https://doi.org/10.1093/nar/gkz601"

Features 

• High throughput scores computation (see fig 2-3)

• Scores can be computed on a basic computer or a supercomputer as well (see benchmark fig 2 and 3)

• Work out of the box : no installation for the scorer and the merger developed using a high performance compiled programming language.

• Conda environment  and video for easy installation of the Viewer/Combiner

• No complex command lines : copy the data and start the software

• Already automated for thousands of pathways 

• Ready to use pathway database with >30,000 human pathways

• Easy use of custom pathways (see doc)

• Can distinguish inhibitor/activator genes in custom pathways (see doc)

• No data alteration : the user can use the latest normalization/noise reduction algorithm during data pre-processing

• Scores manually validated and heavily tested by experienced biologists/immunologists on a large amount of data using antibodies, purified cells. (see publication supp data)

• Step-by-step documentation of 17 pages, continuously improved to cover the user's questions. 

Figure 1.  KEGG Signature "Antigen Processing and Presentation" (89 genes) displayed on a t-SNE plot of 13,500 human peripheral blood mononuclear cells by Single-Cell Signature Viewer .

Figure 2. In this benchmark, Single Cell Signature Scorer computes the scores of 2700 signatures for 1 million of cells in less than 30 min and consumes less than 4 Go RAM.

Figure 4. Single-Cell Signature Explorer is compatible with Single-Cell Spatial Explorer. For an easier coordinates to pixel position conversion you can use Spatial Background Builder .

change log :

2025-07-30 :  new !  Major update of the signature Scorer,  memory usage and speed have been greatly improved (see Benchmarks and doc)

2024-06-13 : Compilation of the merger and scorers with go 1.22.4 for linux, windows, macx86 and macArm64 : speed improvement 7-8%

2023-10-26 : new !  Major update of  LowMemoryScorer : result temp files replaced by a high speed database. Scorer compiled with go 1.21.3. Documentation update.

2023-10-25 :  Spatial Background Builder , a new file conversion tool to use Single-Cell Signature Explorer scRNAseq data with Single-Cell Spatial Explorer.

2023-10-05 : Scorer : memory management,  thread optimisation, compilation with Go 1.21.1 : speed increase : 2%, RAM improvement : 32%. ScorerLowMem : compilation with Go 1.21.1 

2023-04-07 : new !  Mamba/Conda environment for easy installation of Single-Cell Signature Viewer/combiner. Documentation update.

2023-02-03 : scScorer and LowMemoryScorer :  have been updated and compiled with Go1.20 : speed gain : 2-2.5x. Documentation improved.

2022-09-09 : new !  scViewer :  has been updated. It is now possible to adjust dot size.

2022-04-07 : new !  scViewer : "Sort dots" function is now implemented in PNG/SVG export.

2022-04-06 : new !  scViewer : a "Sort dots" menu orders the dots to bring higher values in foreground/background.  Documentation improved.

2022-01-12 : scViewer : improvement of slider bounds calculation.  scViewer & scCombiner : color legend title bug fix. Progress bar for the save image buttons.  documentation improved

2021-12-08 :  documentation improved

2021-11-09 :  documentation improved

2021-10-08 :  we had to remove the merger exe file for windows from the ZIP archive, because it is considered as a virus by google drive. We are sorry for that, the merger windows exe file have to be compiled from source or it is available upon request.

2021-10-06 : Viewer & Combiner : improvement of drop down lists with the picker Input widget : contributor Rüçhan EKREN. documentation improved.

2021-09-29 : Viewer & Combiner : automatic package install. Scorer & Merger : compiled with Go 1.17.1 should be faster. New : Apple Mac version (done by cross compilation. No support since I have no Mac). Documentation improved : how to install software and packages to use the Viewer & Combiner on window is now available on youtube. Please note that this video covers the installation of R, Rstudio, Rtools, that are not supported by us.

2021-06-21 : windows version of  Merger (considered as a virus by google drive in the past) is now available in the zip archive

2021-05-10 : Database update

2021-04-13 : documentation improved

2020-11-25 : Database update

2020-11-16 : CPU management has been changed in the Scorer Low Memory (solve problems of CPU attribution when more than one table is processed). The documentation was improved.

2020-10-19 : Merger update : user lists are now compatible with windows end of lines.

2020-08-27 : Scorer update : a new Scorer with low memory usage has been developed. This new version should be able to process a huge number of cells with limited RAM usage (around 16Go).  Read carefully the documentation,  new  tuning options are available in this version.

2020-05-19 : Databases update

2020-04-01 : update manual : how to compile Go software. A faster progress bar is now used in the scorer and the merger.

2020-03-13 : update manual :  how to solve troubleshooting with some versions of shiny

2020-02-28: Single-Cell Signature Explorer v3.5 : Speed increase of 15% for the scorer compiled with Go 1.14

Manual : 

A detailed manual is available for download in the software zip file  : Manual_Single_Cell_Signature_Explorer.pdf)

Signature Databases compatible with Single-Cell Signature Explorer are available here.

Known issues

1- In some versions of shiny R package (used by the viewer and the combiner) the following error can occur : Listening on http://127.0.0.1:5405 Error in utils::browseURL(appUrl) : ’browser’ must be a non-empty character string.

Solution :

in R session specify the web browser name with the command :

options(browser = 'firefox') 

or 

options(browser = ”firefox”)  (in some version simple quotes are required, in other double quotes)

and restart the viewer with :

source("Run.R")

if the previous solution fails, then a non interactive session must be used, in a terminal type :

 Rscript Run.R

and then open the URL manually in the web browser

2- The maximum number of item that a drop down list can display is limited. This limit is dependent on the browser, and browser version.  For example in Firefox a large list can disappears.  In this case, try to use google chrome or split the table. This can be done using some tools in this website, for example ”Select columns in table”.

3- In some linux distributions the Cairo R pakcage, used by the viewer/combiner, is difficult to install. A solution in this post . We recommend to use the conda environment available in the software zip archive to install the needed R packages automatically. 

Most pre-compiled exe files are now considered as a virus by google drive. I asked  google to validate the binary files without success. 

Meanwhile, a full archive, containing source code, documentation and binary files, is available by clicking on the blue  button bellow with the password :  SignatureExplorer

If any problems,  pre-compiled binary files are available upon request.