analyzeFI: Functional interaction analysis of AS events

View source: R/analyzeFI.R

analyzeFIR Documentation

Functional interaction analysis of AS events

Description

Analyze functional interactions of AS events using Discriminative Random Walk with Restart (DRaWR). It runs a DRaWR on a heterogeneous network containing genes, AS events, and pathways. It then performs GSEA on gene sets related to query genes.

Usage

analyzeFI(object, query, expr, ppi = NULL, pathways = NULL,
    restart = 0.7, num.folds = 5, num.feats = 100, low.expr = 1,
    low.var = NULL, prop.na = 0.05, prop.extreme = 1,
    cor.threshold = 0.3)

Arguments

object

Object of class ASpediaFI

query

a character vector or a data frame containing query genes

expr

a SummarizedExperiment object or matrix containing gene expression profiles (FPKM)

ppi

an igraph object containing known interactions between genes. If NULL, an igraph object containing human gene-gene interactions will be used.

pathways

a GMT file or a named list of pathway gene sets. If NULL, a combined list of HALLMARK, KEGG, and REACTOME pathway gene sets will be used.

restart

a restart probability

num.folds

the number of folds for cross-validation

num.feats

the number of feature nodes to be retained in the final subnetwork

low.expr

Genes with mean expression below low.expr are excluded. AS events for corresponding genes are also excluded.

low.var

AS events with variance below low.var are excluded. If NULL, top 10,000 variable events are used for analysis.

prop.na

AS events with the higher proportion of missing values than prop.na are excluded.

prop.extreme

AS events with the higher proportion of extreme values (0 or 1) than prop.extreme are excluded.

cor.threshold

a pair of AS event and gene with Spearman's correlation greather than cor.threshold are connected in a heterogeneous network.

Value

ASpediaFI object with results of functional interaction analysis

References

Blatti, C. et al. (2016). Characterizing gene sets using discrminative random walks with restart on heterogeneous biological networks. Bioinformatics, 32.

Examples

library(limma)
data(GSE114922.fpkm)
data(GSE114922.psi)
design <- cbind(WT = 1, MvsW = colData(GSE114922.psi)$condition == 'MUT')
fit <- lmFit(log2(GSE114922.fpkm + 1), design = design)
fit <- eBayes(fit, trend = TRUE)
tt <- topTable(fit, number = Inf, coef = 'MvsW')
query <- rownames(tt[tt$logFC > 1 & tt$P.Value < 0.1, ])
head(query)
## Not run: 
GSE114922.ASpediaFI <- analyzeFI(
    GSE114922.ASpediaFI, query,
    GSE114922.fpkm
)

## End(Not run)

nachoryu/ASpediaFI documentation built on Aug. 19, 2022, 7:51 a.m.