multiOutput: Downstream visualization tools

In CAMO-R/Rpackage: Perform cross-species resemblance analysis

Downstream visualization tools

Description

Downstream visualization tools: visualization outputs including overall pathway clustering and output for each pathway The multiOutput is function to visualize pathway level c-scores and d-scoresincluding pathway clustering results with co-membership heatmaps, within-pathway MDS plot (on studies), within-pathway clustering heatmap (on studies), within-pathway posterior DE heatmap, KEGG and Reactome topology plots for each pathway.

Usage

multiOutput(
  mcmc.merge.list,
  dataset.names,
  select.pathway.list,
  ACS_ADS_pathway,
  output = c("clustPathway", "mdsModel", "clustModel", "genePM", "keggView",
    "reactomeView"),
  optK = NULL,
  sil_cut = 0.1,
  use_ADS = FALSE,
  hashtb = NULL,
  keywords_cut = 0.05,
  text.permutation = "all",
  comemberProb_cut = 0.7,
  ViewPairSelect = NULL,
  kegg.species = "hsa",
  KEGG.dataGisEntrezID = FALSE,
  KEGG.dataG2EntrezID = NULL,
  KEGG.pathID2name = NULL,
  reactome.species = "HSA",
  Reactome.dataG2TopologyGtype = NULL,
  Reactome.pathID2name = NULL
)

Arguments

mcmc.merge.list:	a list of merged MCMC output matrices.
dataset.names:	a vector of dataset names matched with the mcmc.merge.list.
select.pathway.list:	a list of selected pathways (containing gene components) for clustering/visualizations.
ACS_ADS_pathway:	a list of four data frames: pathway-specific c-scores, pathway-specific d-scores and their permuted p-value (each row is a pathway and each column is a study).
output:	choose from: "clustPathway" (pathway clustering), "mdsModel"(within-pathway MDS plot on studies), "clustModel" (within-pathway clustering heatmap), "genePM" (within-pathway posterior DE heatmap), "keggView" (KEGG pathway topology, default is human - hsa, for other species, KEGG organism name and gene Entrez ID needs to be provided as 'KEGG.dataG2EntrezID'), "reactomeView" (Reactome pathway topology, default is human - HSA, for other species, Reactome organism name needs to be provided as "reactome.species"). Clustering analysis is not applicable when the number of studies is smaller than 3. "output" cannot be empty.
optK:	Optimal number of clusters. For "clustPathway" output only.
sil_cut:	silhouette index to control scatterness. Larger value indicates tigher cluster and more scattered pathways.
use_ADS:	whether use d-scores for clustering/visualizations. Default is FALSE.
hashtb:	a flat noun-pathway table for text mining. Prepared tables from KEGG and Reactome pathway descriptions for 5 species "hsa", "mmu", "rno", "cel" and "dme"are provided by data(hashtb_hsa), data(hashtb_mmu), data(hashtb_rno), data(hashtb_cel) and data(hashtb_dme). Please refer to Zeng, Xiangrui, et al. "Comparative Pathway Integrator: a framework of meta-analytic integration of multiple transcriptomic studies for consensual and differential pathway analysis." Genes 11.6 (2020): 696.
keywords_cut:	keywords above this cut will be shown in the text mining spreadsheet output.
text.permutation:	select from "all" or "enriched". In text mining, "all" permutates pathways from full pathway.list provided while "enriched" permutates from selected pathways. "all" is suitable for cross-species comparision while "enriched" is recommended for within-species comparision.
comemberProb_cut:	probability below this cut will be colored blue in comembership heatmaps.
ViewPairSelect:	which two datasets to view in the KEGG/Reactome topology plot. All pairs will be considered under default (may take a while).
kegg.species:	KEGG species abbreviation. For "keggView" only. Default is "hsa".
KEGG.dataGisEntrezID:	whether gene names in data are EntrezID. Default is FALSE.
KEGG.dataG2EntrezID:	a data frame which maps gene names in mcmc.merge.list (first column) to Entrez IDs (second column). If NULL & KEGG.dataGisEntrezID=F & kegg.species is one of "hsa", "mmu", "rno", "cel" or "dme", gene symbols will be automatically mapped to EntrezID by Bioconductor packages "org.Hs.eg.db", "org.Mm.eg.db", "org.Rn.eg.db", "org.Ce.eg.db" or org.Dm.eg.db". For "keggView" only.
KEGG.pathID2name:	a list where each element is a KEGG pathway name with correponding pathway ID as its name. ID will be retrieved from KEGGREST if this is NULL.
reactome.species:	Reactome species abbreviation. For "reactomeView" only. Default is "HSA".
Reactome.dataG2TopologyGtype:	a data frame which maps gene names in mcmc.merge.list (first column) and select.pathway.list to gene name types in Reactome topology (second column). For "reactomeView" only.
Reactome.pathID2name:	a list where each element is a Reactome pathway name with correponding pathway ID as its name. ID will be retrieved from reactome.db if this is NULL.

Value

all figures and tables are stored in created folders in the current directory.

Examples

## Not run: 
#mcmc.merge.list from the merge step (see the example in function 'merge')
#select.pathway.list from the pathSelect step (see the example in function 'pathSelect')
#ACS_ADS_pathway from the multi_ACS_ADS_pathway step (see example in 'multi_ACS_ADS_pathway')
data(hashtb_hsa) #include hashtb & pathways for text mining
dataset.names = c("hb","hs","ht","ha","hi","hl",
                  "mb","ms","mt","ma","mi","ml")
#1. step1: select K by elbow plot from consensus clustering
ACSpvalue.mat = ACS_ADS_pathway$ACSpvalue.mat
results = ConsensusClusterPlus(d=t(-log10(ACSpvalue.mat)),maxK=10,reps=50,pItem=0.8,
pFeature=1,title="Consensus Clustering",clusterAlg="hc",innerLinkage="ward.D2",
finalLinkage="ward.D2",seed=12345,plot="png")

#2. step2: run multiOutput with pre-selected K=4
multiOutput(mcmc.merge.list,dataset.names,select.pathway.list,ACS_ADS_pathway,
           output=c("clustPathway","mdsModel","clustModel","genePM","keggView"),
           hashtb=hashtb,optK = 4,keywords_cut=0.2,comemberProb_cut=0.6)

## End(Not run)

CAMO-R/Rpackage documentation built on July 20, 2023, 6:04 a.m.