Help :: Test dataset :: About
The Protein Topology Deviser is an R package for the automatic generation of protein topology code in JSON format. The code can be easily rendered into a graph, by Pfam's custom-domains generator. This excellent tool is not developed by us, but is embedded with the program for users' convenience. A table summary is also prepared.
This is the R package of ProToDeviseR, which can be installed locally. If you are looking for the online version of the program (which requires no installation), visit https://matrinet.shinyapps.io/ProToDeviser/.
This work, as well as, the publication open access (OA) has been supported by:
ProToDeviseR features a fully functional graphical user interface (UI), implemented in R Shiny:
As an input, you can use:
- A UniProt identifier or an NCBI GenPept identifier.
- Raw results from several online resources for protein topology prediction
- A user-prepared table of protein topology annotations.
As an output, you get:
- JSON code of protein features, ready to be processed into graphics in the Image generator tab.
- A table summary of protein features.
The protein features are classified as regions, motifs and markups (Figure 1). When searching with a database accession number (ID), all three are enabled by default:
- Regions: structured domains, repeats or other relatively long (usually, but not always), functional parts of the protein.
- Motifs: short liner motifs, disordered protein parts, signal peptides and transmembrane parts.
- Markups: single-amino acid targes, such as glycosylation, phosphorylation, active or binding site, etc.
Different colour palettes for regions (domains) are provided, for users' preference.
Figure 1. Automatic annotation of CD45 protein topology. Identifier P08575 was searched against UniProt by ProToDeviseR. Table and JSON code were automatically generated and graphic was rendered in the Image generator tab. Regions, motifs and markups are indicated.
ProToDeviser and its R dependencies can be installed as shown in the code box below. Depending on your system, you may need to also install system dependencies (outside R) required by them.
- Click here for external Linux deps, and system-wide installation.
# from CRAN
install.packages("jsonlite")
install.packages("seqinr")
install.packages("dplyr")
install.packages("openxlsx")
install.packages("rentrez")
install.packages("shiny")
install.packages("shinyBS")
install.packages("rclipboard")
install.packages("viridis")
install.packages("DT")
install.packages("textshaping")
install.packages("ragg")
install.packages("markdown")
install.packages("pkgdown")
install.packages("gggenomes")
install.packages("devtools")
install.packages("shinybusy")
# from BioConductor
install.packages("BiocManager")
BiocManager::install("IRanges")
# finally, install ProToDeviseR itself
devtools::install_github("izzilab/protodeviser")Load the library and run the UI function. To use the app graphically, that's all you need. The Help tab provides extensive documentation on the user intefrace, including examples.
library("protodeviser")
protodeviser_ui()ProToDeviser uses the following command-line functions to generate a JSON topology scheme:
id.JSON: for a protein database identifier (ID).predicted.JSON: from predicted features for your protein.custom.JSON: from a (manually prepared) table of features for your protein.
To generate/output features as a table:
json.table: from JSON input of protein features.
ProToDeviser ships a modified version of the domains generator. The official custom domains generator has been decomissioned, but the Domain GFX "playground" demo is an alternative place.
If you use ProToDeviseR in your work, please cite us, as well as Pfam, where the Image Generator tool is from:
- Petrov, P., Izzi, V. ProToDeviseR: an automated protein topology scheme generator. BMC Bioinformatics 26, 71 (2025). https://doi.org/10.1186/s12859-025-06088-2
- Mistry J, Chuguransky S, Williams L, Qureshi M, Salazar GA, Sonnhammer ELL, et al. Pfam: the protein families database in 2021. Nucl Acids Res. 2021;49:D412–9.