Universal concept signature analysis: genome-wide quantification of new biological and pathological functions of genes and pathways

Abstract

Identifying new gene functions and pathways underlying diseases and biological processes are major challenges in genomics research. Particularly, most methods for interpreting the pathways characteristic of an experimental gene list defined by genomic data are limited by their dependence on assessing the overlapping genes or their interactome topology, which cannot account for the variety of functional relations. This is particularly problematic for pathway discovery from single-cell genomics with low gene coverage or interpreting complex pathway changes such as during change of cell states. Here, we exploited the comprehensive sets of molecular concepts that combine ontologies, pathways, interactions and domains to help inform the functional relations. We first developed a universal concept signature (uniConSig) analysis for genome-wide quantification of new gene functions underlying biological or pathological processes based on the signature molecular concepts computed from known functional gene lists. We then further developed a novel concept signature enrichment analysis (CSEA) for deep functional assessment of the pathways enriched in an experimental gene list. This method is grounded on the framework of shared concept signatures between gene sets at multiple functional levels, thus overcoming the limitations of the current methods. Through meta-analysis of transcriptomic data sets of cancer cell line models and single hematopoietic stem cells, we demonstrate the broad applications of CSEA on pathway discovery from gene expression and single-cell transcriptomic data sets for genetic perturbations and change of cell states, which complements the current modalities. The R modules for uniConSig analysis and CSEA are available through https://github.com/wangxlab/uniConSig.

An integral genomic signature approach for tailored cancer targeted therapy using genome-wide sequencing data

“…According to literature, the AUC and Act Area are much better quantifiers of drug responses than IC50 6 . To uniform multi-OMIC features, we formulated a Genomic-feature Matrix Transposed (GMT) format for compiling binary multi-OMIC features, similar to that used for compiling gene concepts 7,8 . Using this format, we analyzed the expression profiling data and exome sequencing data from GDSC and compiled an integrated dataset combining the genomic features including upregulated genes, downregulated genes, mutated genes, and mutation hotspots.…”

“…This can be achieved by extracting the genes contributing to the resistance related genomic features in our GDSC iGenSig model. The resulting resistance gene list can be then used to explore the enriched pathways based on the concept signature enrichment analysis (CSEA) developed in our previous study, which is designed for deep functional assessment of the pathways enriched in an experimental gene list 8 . Our result showed that the most significantly upregulated pathways characteristic of Erlotinib resistance signature include MTORC1 signaling and E2F target gene signature (Figure 5c, Supplementary Fig.…”

“…The negative contributing genes are defined as downregulated genes or mutated genes without mutation hotspots. The pathways enriched in the positive or negative contributing genes for Erlotinib resistance are analyzed by the Concept Signature Enrichment Analysis (CSEA) developed in our previous study 8 . The resulting top 30 pathways are disambiguated via correcting the crosstalk effects between pathways, to reveal independent pathway modules 28 .…”

“…A p-value <0.01 is used as cutoff for disambiguation. The functional associations between the significant pathways are then assessed using our CSEA method as we previously described 8 , and the CSEA scores are then scaled between -1 and 1 and visualized using correlogram. The pathway network was visualized using the 'igraph' R package (ver.…”

“…In particular, our iGenSig models demonstrated outstanding performances in crossing datasets and tolerating the bias in the detected genomic features. IGenSig models can be managed in every detailed step, and the underlying pathways can be biologically interpreted through the concept signature enrichment analysis we developed8 . The performance of iGenSig models appears to at least in part depend on the availability of outstanding responders and significant genomic correlates.…”