The Signaling Pathways Project, an integrated ‘omics knowledgebase for mammalian cellular signaling pathways

Scott A. Ochsner, David Abraham, Kirt Martin, Wei Ding, Apollo McOwiti, Wasula H. Kankanamge, Zichen Wang, Kaitlyn J. Andreano, Ross A. Hamilton, Yue Chen, Angelica Hamilton, Marin L. Gantner, Michael Dehart, Shijing Qu, Susan G. Hilsenbeck, Lauren B. Becnel, Dave Bridges, Avi Ma’ayan, Janice M. Huss, Fabio Stossi, Charles E. Foulds, Anastasia Kralli, Donald P. McDonnell, Neil J. McKenna

TL;DR

The Signaling Pathways Project (SPP) is a new web-based knowledgebase that integrates vast amounts of public transcriptomic and ChIP-Seq data to help researchers understand mammalian cellular signaling pathways. It introduces "consensomes," a novel meta-analysis technique to predict gene regulatory relationships, which has been validated through literature and experimental studies.

What Problem Does This Paper Address?

Understanding how cells respond to signals is fundamental to biology and medicine. This involves intricate networks of proteins called signaling pathways. While a wealth of "omics" data (like gene expression and protein binding) exists from experiments, it's often siloed, difficult to access, and underutilized. Researchers struggle to connect specific pathway components to their downstream effects on gene regulation, especially for less-studied pathways. This limits hypothesis generation and the discovery of new therapeutic targets. The SPP aims to solve this by creating a unified, searchable resource that makes this data FAIR (Findable, Accessible, Interoperable, and Reusable).

Methods

The core of the SPP is its biocuration pipeline and the development of "consensomes."

  • Biocuration: Public transcriptomic and ChIP-Seq datasets were systematically collected, processed, and annotated. Crucially, experiments were mapped to specific signaling pathway nodes (receptors, enzymes, transcription factors, co-nodes) and biosamples.
  • Consensomes: This is the key innovation. Instead of analyzing individual experiments, SPP performs a meta-analysis across many experiments related to a specific pathway node or family. This generates a "consensus signature" that ranks genes based on their consistent differential expression (transcriptomics) or promoter occupancy (ChIP-Seq) across these experiments. This allows for the prediction of likely gene targets for a given pathway node.
  • Web Interface (Ominer): A user-friendly interface allows researchers to query the knowledgebase using single genes, Gene Ontology terms, or custom gene lists, and to explore the generated consensomes and "Regulation Reports."

Datasets

The SPP integrates over 10,000 public transcriptomic and ChIP-Seq experiments. These datasets are sourced from public repositories like the Gene Expression Omnibus (GEO) and Sequence Read Archive (SRA). The paper does not introduce new benchmark datasets but rather organizes and annotates existing ones.

Key Insights and Ideas

  • The Power of Meta-Analysis for Discovery: By aggregating data from numerous experiments, SPP can identify robust, consistent patterns that might be missed in individual studies. This is particularly useful for uncovering less-obvious regulatory relationships.
  • "Consensomes" as Predictive Tools: The concept of a consensome is powerful because it moves beyond individual experimental results to a community-wide consensus on a node's likely targets. This provides a data-driven starting point for hypothesis generation.
  • Bridging the Gap Between 'Omics and Bench Science: SPP is designed to be directly useful to bench researchers, providing a platform to explore complex 'omics data without requiring extensive bioinformatics expertise. The validation studies highlight how SPP can predict novel relationships that can then be experimentally confirmed.
  • FAIR Data Principles in Action: The project emphasizes making data Findable, Accessible, Interoperable, and Reusable, with features like DOIs for datasets and a public web interface.

Key Results

  • Consensome Validation: 90% of highly ranked targets in nuclear receptor consensomes were previously identified in the literature. Furthermore, 82% of these validated targets were also highly ranked in corresponding ChIP-Seq consensomes, demonstrating cross-validation between different data types.
  • Discovery of Novel Relationships: Bench experiments confirmed previously uncharacterized node-gene target regulatory relationships predicted by SPP, such as the role of TPD52L1 in estrogen receptor signaling and MBOAT2 in androgen receptor signaling.
  • Enrichment of Metabolic Regulators: In the mouse liver, genes highly ranked in the "all nodes" transcriptomic consensome were significantly enriched for metabolic enzymes, with a notable proportion being rate-limiting enzymes, suggesting SPP can identify key regulators of biological processes.

Limitations and Future Work

  • Data Type Limitations: Currently, SPP primarily focuses on transcriptomic and ChIP-Seq data. The authors acknowledge that incorporating metabolomic and proteomic data in future versions would provide a more comprehensive view of signaling pathway effects.
  • Dataset Bias: The knowledgebase reflects biases in publicly archived datasets, meaning some nodes or biosamples may be better represented than others.
  • Experimental Context: The ranking of targets in consensomes can be influenced by specific experimental conditions. If a regulatory relationship only occurs under very specific circumstances not captured in the available data, it might not be highly ranked.
  • Future Work: The authors envision expanding SPP to include other 'omics data types and improving its ability to handle more complex experimental designs. They also aim to enhance its utility for hypothesis generation and data validation for bench researchers.

Summary updated

Abstract

Mining of integrated public transcriptomic and ChIP-Seq (cistromic) datasets can illuminate functions of mammalian cellular signaling pathways not yet explored in the research literature. Here, we designed a web knowledgebase, the Signaling Pathways Project (SPP), which incorporates community classifications of signaling pathway nodes (receptors, enzymes, transcription factors and co-nodes) and their cognate bioactive small molecules. We then mapped over 10,000 public transcriptomic or cistromic experiments to their pathway node or biosample of study. To enable prediction of pathway node-gene target transcriptional regulatory relationships through SPP, we generated consensus 'omics signatures, or consensomes, which ranked genes based on measures of their significant differential expression or promoter occupancy across transcriptomic or cistromic experiments mapped to a specific node family. Consensomes were validated using alignment with canonical literature knowledge, gene target-level integration of transcriptomic and cistromic data points, and in bench experiments confirming previously uncharacterized node-gene target regulatory relationships. To expose the SPP knowledgebase to researchers, a web browser interface was designed that accommodates numerous routine data mining strategies. SPP is freely accessible at https://www.signalingpathways.org .
BibTeX
@inproceedings{ochsner2019signaling,
  author = {Scott A. Ochsner and David Abraham and Kirt Martin and Wei Ding and Apollo McOwiti and Wasula H. Kankanamge and Zichen Wang and Kaitlyn J. Andreano and Ross A. Hamilton and Yue Chen and Angelica Hamilton and Marin L. Gantner and Michael Dehart and Shijing Qu and Susan G. Hilsenbeck and Lauren B. Becnel and Dave Bridges and Avi Ma’ayan and Janice M. Huss and Fabio Stossi and Charles E. Foulds and Anastasia Kralli and Donald P. McDonnell and Neil J. McKenna},
  title = {{The Signaling Pathways Project, an integrated ‘omics knowledgebase for mammalian cellular signaling pathways}},
  booktitle = {Scientific},
  year = {2019},
  doi = {10.1038/s41597-019-0193-4},
}
Figure