Assessing the impact of mutations found in next generation sequencing data over human signaling pathways

Abstract

Modern sequencing technologies produce increasingly detailed data on genomic variation. However, conventional methods for relating either individual variants or mutated genes to phenotypes present known limitations given the complex, multigenic nature of many diseases or traits. Here we present PATHiVar, a web-based tool that integrates genomic variation data with gene expression tissue information. PATHiVar constitutes a new generation of genomic data analysis methods that allow studying variants found in next generation sequencing experiment in the context of signaling pathways. Simple Boolean models of pathways provide detailed descriptions of the impact of mutations in cell functionality so as, recurrences in functionality failures can easily be related to diseases, even if they are produced by mutations in different genes. Patterns of changes in signal transmission circuits, often unpredictable from individual genes mutated, correspond to patterns of affected functionalities that can be related to complex traits such as disease progression, drug response, etc. PATHiVar is available at: http://pathivar.babelomics.org.

Figure 1.

Schema of the analysis of impact of mutations over signaling pathways. (A) VCF files are uploaded in the system. (B) KEGG pathway definitions are selected. (C) VCF files are annotated using information contained in CellBase. Consequence type, SIFT, PolyPhen and phastCons indexes are associated to each variant position in the VCF. (D) Tissue is selected (or user-defined pattern of gene presence/absence is uploaded) and the unperturbed map of signal transduction (corresponding to the functional genes in the tissue) is deduced from the presence/absence of the genes in the pathway. (E) Depending on the inheritance pattern (dominant/recessiv/compound heterozygote) the expressed but damaged proteins are removed from the model and the net signal transduction is inferred again which produces (F) the perturbed map of signal transduction. The differences between the unperturbed (D) and the perturbed (F) signal transduction maps are reported by PathiVar. The bottom right of the figure show the symbols used to denote expressed and non-expressed genes, genes harboring deleterious mutations and the interactions that produce signal transduction.

Figure 2.

Output of a fictitious combination of gene expression and mutations in version of the Calcium signaling pathway modified for illustrating the examples. The figure illustrates the possible effects due to the combination of gene expression and gene loss of function within the topology of a pathway. In the figure, blue background indicates gene expression, while gray background means no expression in the tissue studied. Yellow background means unknown expression in the tissue. Black arrow indicates signal transmission whereas gray arrow means no signal is transmitted. (A) signal deactivation: the gene GNA11 is expressed but harbors a deleterious mutation and the signal does not flux downstream; (B) signal activation: the repressor INH harbors a deleterious mutation and therefore cannot inhibit PLN and the signal flux, that would be interrupted here with a functional protein, is activated instead; (C) neutral effect: PLCG1 with a deleterious mutation does not transmit the signal, however, The signal is transmitted anyway from EGFR to both, ITPR1 and PRKCA, through the protein PLCD3, because the receptor-effector signaling circuit is internally redundant; (D) neutral effect: several examples show how mutations affect to genes that are not expressed in the tissue studied (CACNA1A, CACNA1I, RYR1, etc.) and consequently have no effect in the particular tissue of study.