In sandraTL/PathQuant: Annotation of gene-metabolite pairs using topology of KEGG metabolic pathway maps

Package 'PathQuant'

Version 1.0

PathQuant: A bioinformatic tool to quantitatively annotate gene-metabolite pairs using the KEGG pathway database.

Author Sandra Therrien-Laperrière

Description

PathQuant aims to systematically and quantitatively annotate gene-metabolite pairs throught mapping onto KEGG metabolic pathway maps to reveal biological insight using their topology.

To do so, PathQuant computes shortest reactional distance (srd), which is define as the number of reaction separating a gene (represented as the edge of the pathway) and a metabolite (represented as the node of the pathway).

We developped a tool to able the computation of srd for gene-metabolite pairs.

Info

The example data used is from the Shin & al Study [1]. Their gene-metabolite associations are available in the 'shin' data.frame included in the package as well as 'completeGeneDF' the complete list of reported genes and 'completeMetaboliteDF' the complete list of measured metabolites in the study.

Fonctions

get.srd

Description

This function enables the user to input their list of gene-metabolite pairs and compute srd values using all metabolic pathways of KEGG by default or using a selected list of metabolic pathways in KEGG. Srd values are computed when both the gene and the metabolite of a given pair is mapped within a same metabolic pathway. Note the topology of the used pathways is kept.

The output of the function is a dataframe with annotation of gene product, metabolite class, enzyme EC number, the computed srd as well as the reacional path used for srd copmutation. Note that an infinite distance (Inf) means that both gene and metabolite of a given pair are on the same map but no path exist between them. If a srd has a 'Not Computed' value, it means neither the gene or the metabolite (or both) are on any of the selected pathway.

df <- get.srd(shin, "hsa" ,"All", F, F, F)

gene.distribution

Function plotting the srd values computed by the get.srd function. The plot depicted as frequency bars, which represent the number of pairs with a given srd. Frequency bars are in heatmap colors to reflect the size of the srd values.

srd.distribution(df[,8])

gene.distribution

Function computes srd between a selected gene and a specific list of metabolites within a given metabolic pathway. The plot is depicted as frequency bars, which represent the number of metabolites at a given srd of the selected gene. Frequency bars are shown in grey for metabolites that are not associated with the selected gene and in red if there is at least one metabolite associated with this gene.

gene.distribution("hsa01100", "hsa:1373", shin, completeMetaboDF)
gene.distribution("hsa01100", "hsa:27165", shin, completeMetaboDF)

heatmap

Function that output a heatmap to visualize srd values computed between every gene-metabolite pairs in input using a specified pathway map. Heatmap of srd computed between every gene and metabolite in pairs mapped on the genome-scale metabolism reconstruction of KEGG. Columns represent genes and rows represent metabolites. The calculated distance is shown in each cell with the corresponding color code (from red - closest; to yellow - farthest). Th srd computed for the gene-metabolite associations are reported in cells with a thick black border. Inf = infinite value, which mean that there is no known path between gene and metabolite.

heatmap(shin, "hsa01100", F)

permutationTest

Statistical permutation test function to assess the relevance of associated gene-metabolite pairs vs. randomly selected pairs. This function produces permutation sets (size of real associated data) by creating random gene-metabolite associations and calcuates the median distance of every permutation to output : 1) 'medians' 2) 'histogram' 3) 'pvalue'. Replacing gene : random genes are chosen within the set of measured genes catalysing +/- 1 reaction(s). Gene catalysing 1 reaction are repalce by a gene calatysing 1 reaction from set of measured genes 'completeGeneDF'. Metabolites are replaced by any of the measured metabolites: 'completeMetaboDF'.

p <- permutationTest("hsa01100", shin,
                         completeGeneDF, completeMetaboDF, 1000, "histogram")

1. Shin, So-Youn et al. “An Atlas of Genetic Influences on Human Blood Metabolites.” Nature genetics 46.6 (2014): 543–550. PMC.

sandraTL/PathQuant documentation built on May 29, 2019, 1:45 p.m.