Frequently asked questions

What is enrichment of kinase substrates?

Protein kinases are enzymes that catalyze the transfer of a phosphate group from ATP to another protein on a threonine, a serine, or a tyrosine residue. The reversible addition of a phosphate group to a protein can profoundly affect its reversible activity, stability, localization, and interactions with other molecules. Of the hundreds of thousands of potential phosphorylation sites, each protein kinase recognizes between one to a few hundred substrates. Thus, kinases serve a critical and central role in regulating essentially all cellular processes, and their aberrant activity is recognized as a cause of many human diseases. Phosphoproteomic signatures, which represent the differential phosphorylation state of cellular proteins between two cellular states, can provide a snapshot of the intracellular signaling networks that are differentially active between two conditions, for instance, between diseased and healthy cells. The enrichment of known kinase substrates in a set of differentially phosphorylated proteins can serve as a marker of the activity of the upstream kinases, thus providing insights into physiological and pathophysiological mechanisms.

What is KEA?

Kinase Enrichment Analysis 3 (KEA3) infers upstream kinases whose putative substrates are overrepresented in a user-inputted list of genes or differentially phosphorylated proteins. The KEA3 database contains putative kinase-substrate interactions collected from publicly available datasets. Gene sets of putative kinase substrates are used as the primary units of analysis in KEA3. These gene sets are organized in gene set "libraries." Libraries are supersets of kinase substrate sets that are aggregated based on the database from which they are derived.

How librares are constructed?

All protein and gene symbols were mapped to 2019 HGNC-approved gene symbols using the custom R package genesetr. Gene or protein symbols that could not be mapped using synonym or alias matching were discarded. The set of 520 unique HGNC-mappable human protein kinases used for KEA3 are derived from Manning et al., Miranda-Saavedra and Barton, and the Illuminating the Druggable Genome (IDG) project.

Kinase-substrate interactions libraries

• The Cheng et al. kinase-substrate interaction network (KSIN) dataset incorporates human kinase-substrate interactions from the Phospho.ELM, HPRD, PhosphoNetworks, and PhosphoSitePlus databases.
• PTMsigDB contains a collection of kinase phosphorylation signatures derived from PhosphoSitePlus.
• PhosD predicts kinase-substrate interactions based on protein domains.

Protein-protein interactions libraries

• The BioGRID and MINT databases contain PPIs from high- and low-throughput experiments that were manually curated from the literature.
• Mentha is a PPI network (PPIN) that contains molecular interactions aggregated and updated weekly from MINT, IntAct, BioGRID, MatrixDB, and the Database of Interacting Proteins (DIP).
• HIPPIE aggregates experimentally determined PPIs from IntAct, MINT, BioGRID, HPRD, DIP, BIND, and MMPI.
• Cheng et al. used PPIs collected fromIntAct, MINT, BioGRID, DIP, HPRD, and MIPS MPact.
• PrePPI consists of predicted PPIs that were determined using a Bayesian framework that relies on three-dimensional structural, functional evolutionary, and expression information to make PPI predictions.

Other libraries

STRING KEA3 library includes physical interaction, co-expression, co-occurrence in the literature, and evolutionary co-occurrence, among other association types.

What is kinase substrate over-representation analysis?

The goal of KEA3 is to predict kinases associated with user-input sets of genes/proteins. Discrete query gene/protein sets are compared to KEA3 libraries of putative kinase substrate sets assembled from multiple publically available datasets. The Fisher's Exact Test, with a background size of 20,000, is used to compare the input gene/protein set to the kinase substrate sets in order to determine which kinases may be most closely associated with the input gene/protein set.

What types of input KEA3 accepts?

Users can submit a set of human gene or protein symbols for kinase enrichment analysis. Commonly, the gene set may be a set of differentially expressed genes, or a set of differentially phosphorylated proteins with the goal of forming hypotheses about which upstream kinases might be responsible for the changes in the phenotype.

What does kinase co-expression networks visualization show?

The KEA3 application displays a scatter plot of points representing human kinases based on their co-expression similarity. This kinase coexpression network was constructed by conducting a Weighted Gene Co-expression Network Analysis (WGCNA) over human kinase expression data. WGCNs were constructed from GTEx samples, TCGA samples, and ARCHS4 samples. The network layouts were determined using Allegro Fruchterman-Reingold Clustering in Cytoscape. For the GTEx networks, WGCNA kinase module summary expression profiles, also known as Eigen-genes, were correlated with gene expression profiles from all GTEx tissue samples to determine strong module-tissue associations. Each GTEx WGCN module may be colored by its most highly correlated tissue. Tissues are aggregated into more general tissue types (e.g. brain) and more specific tissue types (e.g. cerebellum, frontal cortex):

TCGA module Eigen-genes were correlated with tumor gene expression profiles to determine strong module-tumor type associations. Each TCGA WGCN module may be colored by its most highly correlated tumor type. Finally, the ARCHS4 module Eigen-genes were correlated with ARCHS4 tissue types. The ARCHS4 network modules may be colored by the most highly correlated tissue type.