runOntoscope.R

# runOntoscope.R
#
# Purpose:  Run an "Ontoscope" query, and analysis based on
#                and extending the mogrify workflow of
#                Rackham et al. (2016).
#
# V 0.00001
# Date:     March 29. 2016
# Author:   Boris Steipe and BCB420 class
#
# V 0.00001 Very first crude code to integrate modules
#
# ==========================================================



setwd("~/Documents/07.TEACHING/50.8-BCB420-JTB2020 2016/BCB420/Ontoscope")

# ==== INITS =============================
source("../Ontoscope/phylify/ontology-explorer.r")



# ========================================




# load the ontology
load("../Ontoscope/phylify/COdat.RData")

mogList <- getMogrifyIDs()

targetParent <- "FF:0000592"


# use gather module to get a list of background Fantom IDs
x1 <- neighborhood(COdat, order=1, nodes= targetParent, mode="out")
x2 <- neighborhood(COdat, order=2, nodes= targetParent, mode="out")
x3 <- neighborhood(COdat, order=3, nodes= targetParent, mode="out")

parent <- "FF:0101581"
bck <- neighborhood(COdat, order=2, nodes=parent, mode="in")

# source Fantom functions
source("./fantom_import/fantom_main.R")
fantom

x <- fantomSearch("cardiac")
cell <- levels(x$FANTOM.5.Ontology.ID)[400]
bck <- levels(x$FANTOM.5.Ontology.ID)[1:2]

request <- c(cell, bck)

fantomOntology(request)

fantomSummarize(5)  #parameter is min. expression count threshold

# load up CONTRAST

source("./contrast/contrast.R")

# NOTE: change format!
counts <- fantomCounts[ , -1]
rownames(counts) <- fantomCounts$short_description
colnames(counts) <- c("fib", "card1", "card2")

diffExp <- contrast_v2(counts)


# Build a network

# Normalize the STRING file by executing functions in ./WEAVE/normalizeWeave.R
# If this has been run before, you only need to reload the file.
# load(file = STRINGnew)

# use only high-confidence edges (score > 900)
tmp <- src
src <- src[src$combined_score > 900, ]
nrow(src)

# build the WEAVE network using WEAVE-STRING.R fucntions
# STRGRAPH <- graph_from_data_frame(src, directed = FALSE)



load("MARA.RData")
# needs code here that builds a network for the cell-types we are working with.
# ....


# using tools from TRRUST_network to load TRRUST GRN
#works if ontoscope is the working directory
#This is the entire TRRUST network with no subsetting
#Directed from Transcription Factor to Gene

source("./TRRUST_network/TRRUST_network.R")
trrust <- loadTRRUST()
setMode(1)
genes <- trrust[,1]
trrust_nodes <- getNodes(trrust)
trrust_edges <- getEdges (trrust)
trrust_nodes <- getWeights(trrust_nodes,trrust_edges)

TRRUST_GRNGRAPH <- graph_from_data_frame(trrust_edges, directed = TRUE)



# [END]