Resources for gene-centric single sample Gene Set Enrichment Analysis (ssGSEA) of gene expression data (e.g. mRNAs, proteins) and site-centric PTM Signature Enrichment Analysis (PTM-SEA) [1] of phosphoproteomics data sets using the PTM signatures database (PTMsigDB) [1].
This is an updated version of the original ssGSEA [2,3] R-implementation. Depending on the input dataset and chosen database (gene sets or PTM signatures), the software performs either ssGSEA or PTM-SEA, respectively. The Molecular Signatures Database (MSigDB) [4] provides a large collection of curated gene sets. Gene sets are stored as plain text in GMT format. A current version of MSigDB gene set collections can be found in the db/msigdb
subfolder. MSigDB gene sets are realeased under Creative Commons Attribution 4.0 International License. The license terms can be found in thedb/msigdb
folder.
File formats supported by ssGSEA2.0/PTM-SEA are Gene Cluster Text GCT v1.2 or GCT v1.3 files. Morpheus provides a convenient way to convert your data tables into GCT format.
For information about the GSEA method and MSigDB please visit http://software.broadinstitute.org/gsea/.
Please check out our new website for PTMsigDB. We have updated PTMsigDB to version v1.9.0 in which we provide better and more consistent annotation of each PTM site. We have also inlcuded a disease category comprising of signatures associated to certain diseases curated from the table Disease-associated_sites
available at PhosphoSitePlus (PSP) [5].
The PTM signatures database (PTMsigDB) is a database comprised of modification site-specific signatures of perturbations, kinase activities and signaling pathways curated from more than 2,500 publications which provides the foundation to perform PTM-SEA. A unique advantage of PTMsigDB over other pathway databases is the annotation of each PTM site with its reported direction of change upon a specific perturbation or signaling event which is incorporated into the scoring scheme of PTM-SEA. The foundation of PTMsigDB is PhosphoSitePlus (PSP) [5], a comprehensive systems biology resource for PTMs, which provides high-quality curation and annotation of PTMs at the individual residue level. A collection of PTM sites, whose levels are collectively regulated in a curated pathway or upon a perturbation, are defined as a signature set. Signature sets in PTMsigDB can be separated into different categories: 1) Perturbation signatures derived from treatment of cells with perturbagens such as small molecules or growth factors; 2) Signature sets of molecular signaling pathways; 3) Kinase-substrate signatures; and 4) Disease-associated signature sets.
To ensure a high degree of compatibility to phosphorylation datasets generated by different software packages and searched against different protein sequence databases, PTMsigDB represents signatures using three different identifiers to represent phosphorylation sites: 1) PSP site group ID; 2) UniProt-centric ID; 3) Flanking sequence (Table 1). While the PSP site group ID provides an unambiguous representation of PTM sites within protein families and across species [5], using this type of identifier restricts the analysis to PTM sites present in PSP. We generally recommend to usinge the flanking sequence as site identifier, since these are more invariant to updates made to protein sequence databases.
Database format | Site accession | Example in PTMsigDB | Example in dataset | Download |
---|---|---|---|---|
UniProt-centric | Uniprot_acc;site-type;direction | Q06609;Y315-p;u | Q06609;Y315-p | human mouse rat |
Flanking sequence | +/-7aa flanking seq-type;direction | ETRICKIYDSPCLPE-p;u | ETRICKIYDSPCLPE-p | human mouse rat |
PSP site group id | site_grp_id-type;direction | 448324-p;u | 448324-p | human mouse rat |
Table 1: PTM site representation in PTMsigDB. The direction of change for a PTM site in a signature is indicated by ;u (up-regulation) or ;d (down-regulation). Please note that the annotation of directionality is a feature of PTMsigDB (column: Example in PTMsigDB) and must not be included when generating compatible site identifier for a particular dataset (column: Example in dataset).
PTM-Signature Enrichment Analysis (PMT-SEA) is a modified version of ssGSEA to perform site-specific signature analysis by scoring PTMsigDB's bi-directional signature-sets. The input to PTM-SEA is a single site-centric data matrix, m, stored in GCT v1.2 or GCT v1.3 format and PTM signatures database (PTMsigDB). Each row in m represents a single phosphorylation site confidently localized to a specific amino acid residue, with measured abundances across samples specified in columns in m. Multiple phosphorylation sites detected on the same peptide have to be converted into separate site-specific entities for every site. While some proteomics software packages, such as MaxQuant [6], readily produce single site-centric PTM reports, the use of other software packages might require additional preprocessing steps.
ssGSEA2.0/PTM-SEA can be run on a local PC/MAC in R or RStudio. In addition, ssGSEA2.0/PTM-SEA can be access on Broad's public GenePattern [7] server. Below we provide instructions how to run ssSGEA2.0/PTM-SEA.
We provide an example dataset that can be used to test PTM-SEA. The dataset is based on Supplemental Table 6 in [1].
Single site-centric phosphoproteome dataset
GenePattern is a powerful platform to deploy and run software or entire analysis pipelines in a web browser [7]. We have implemented ssGSEA2.0/PTM-SEA as GenePattern module which can be accessed at the link below. Please note that access to the public GenePattern server requires a free registration.
PTM-SEA in GenePattern: https://tinyurl.com/PTM-SEA-GP
The script ssgsea-gui.R
requires little or no knowledge of R or on how to use the command line. Input files and databases can be specified via Windows file dialogs that will be automatically invoked. The first dialog lets you choose a folder containing input files in GCT v1.2 or GCT v1.3 format. The script loops over all GCT files in this directory and runs ssGSEA on each file separately. The second dialog window lets the user choose one or multiple gene set databases in GMT format such as MSigDB. A current version of MSigDB databases can be found in the db
subfolder.
To run the script source it into a running R-session.
- RStudio: open the file and press 'Source' in the upper right part of the editor window
- R-GUI: drag and drop this file into an R-GUI window
In order to invoke file dialogs as decribed above, the XQuartz X Window System is required. Once installed ssgsea-gui.R
can be sourced into an R session.
For integration of ssGSEA2.0/PTM-SEA into your own analysis pipelines we recommend to use the ssgsea-cli.R
script which has been successfully tested on Windows, Mac and Linux OS. Please see ssgsea-cli.R --help
for instructions.
Other paramaters for ssGSEA/PTM-SEA can be altered inside the parameters section in ssgsea-gui.R
or as arguments on the command line. The default parameters have been choosen carefully and should provide reliable results for most use-case scenerios.
Original code written by Pablo Tamayo. Adapted with additional modifications by D. R. Mani and Karsten Krug. Adaptions include:
- support of multiple CPU cores (
doParallel
R-package) - support of GCT v1.3 format using functions from cmapR
- improved handling of missing values
- scoring of directional gene sets (PTMsigDB)
- basic error handling
- improvements in runtime performance
- additional output files like rank plots and parameter files
License Agreement for MSigDB v6.0 and above can be found here.
-
Krug, K., Mertins, P., Zhang, B., Hornbeck, P., Raju, R., Ahmad, R., . Szucs, M., Mundt, F., Forestier, D., Jane-Valbuena, J., Keshishian, H., Gillette, M. A., Tamayo, P., Mesirov, J. P., Jaffe, J. D., Carr, S. A., Mani, D. R. (2019). A curated resource for phosphosite-specific signature analysis, Molecular & Cellular Proteomics (in Press). http://doi.org/10.1074/mcp.TIR118.000943
-
Barbie, D. A., Tamayo, P., Boehm, J. S., Kim, S. Y., Susan, E., Dunn, I. F., . Hahn, W. C. (2010). Systematic RNA interference reveals that oncogenic KRAS- driven cancers require TBK1, Nature, 462(7269), 108-112. https://doi.org/10.1038/nature08460
-
Abazeed, M. E., Adams, D. J., Hurov, K. E., Tamayo, P., Creighton, C. J., Sonkin, D., et al. (2013). Integrative Radiogenomic Profiling of Squamous Cell Lung Cancer. Cancer Research, 73(20), 6289-6298. http://doi.org/10.1158/0008-5472.CAN-13-1616
-
Subramanian, A., Tamayo, P., Mootha, V. K., Mukherjee, S., Ebert, B. L., Gillette, M. A., et al. (2005). Gene set enrichment analysis: a knowledge-based approach for interpreting genome-wide expression profiles. Proceedings of the National Academy of Sciences of the United States of America, 102(43), 15545-15550. http://doi.org/10.1073/pnas.0506580102
-
Hornbeck, P. V., Zhang, B., Murray, B., Kornhauser, J. M., Latham, V., & Skrzypek, E. (2015). PhosphoSitePlus, 2014: mutations, PTMs and recalibrations. Nucleic Acids Research, 43(D1), D512-D520. https://doi.org/10.1093/nar/gku1267
-
Cox, J., & Mann, M. (2008). MaxQuant enables high peptide identification rates, individualized p.p.b.-range mass accuracies and proteome-wide protein quantification. Nature Biotechnology, 26(12), 1367-1372. https://doi.org/10.1038/nbt.1511
-
Reich, M., Liefeld, T., Gould, J., Lerner, J., Tamayo, P., & Mesirov, J. P. (2006). GenePattern 2.0. Nature Genetics, 38(5), 500-501. https://doi.org/10.1038/ng0506-500