cas9OCT4_hsZygote_diffexDT: Differential expression from RNAseq assay of OCT4(POU5F1)...

In adamsardar/metaDEGth: Combining differential expression P-values in a statistically consistent fashion whilst respecting background signal/noise trends

Differential expression from RNAseq assay of OCT4(POU5F1) CRISPR-Cas9 knockouts in human zygote

Description

GSE100118 is a CRISPR knockdown study of the pluripotency transcription factor OCT4 during human embryogenesis.

Generated 10 December 2019 from the trimmed reads associated with project on GEO. These were downloaded and aligned against the human genome using Kallisto before transcript abundances were estimated using limma-voom. Finally, differential gene expression values were computed.

Usage

data(cas9OCT4_hsZygote_diffexDT)

Format

A data.table with 15,028 rows, documenting differential gene expression (logFC and unadjusted P-values) of 15,003 ENSEMBL gene entities in zygotes grown with and wihout the OCT4 transcription factor gene. Addition gene symbols, entrez gene IDs and uniprot annotations also included (duplicated ENSG entries result from many-to-one mappings)

Source

https://www.ncbi.nlm.nih.gov/geo/query/acc.cgi?acc=GSE100118

References

Fogarty NME, McCarthy A, Snijders KE, Powell BE et al. Genome editing reveals a role for OCT4 in human embryogenesis. Nature 2017

Pimentel H, Bray NL, Puente S, Melsted P, Pachter L. Differential analysis of RNA-seq incorporating quantification uncertainty. Nat Methods. 2017

Law CW, Chen Y, Shi W, Smyth GK. voom: Precision weights unlock linear model analysis tools for RNA-seq read counts. Genome Biol. 2014

Examples

library(ggplot2)
library(magrittr)
data(cas9OCT4_hsZygote_diffexDT)

cas9OCT4_hsZygote_diffexDT %>% ggplot(aes(x = koOCT4_logFC, y = -log10(koOCT4_pValue))) + 
    geom_point(aes(colour = p.adjust(koOCT4_pValue, "fdr") < 0.01)) +
    scale_colour_manual(values = c("darkgrey","dodgerblue")) +
    theme_bw()
    
   

cas9OCT4_hsZygote_diffexDT[,qplot(koOCT4_pValue, bins = 100)]

OCT4knockout_betaUniformModel <- fitBetaUniformMixtureDistribution(cas9OCT4_hsZygote_diffexDT$koOCT4_pValue, nStarts = 20)


OCT4knockout_betaUniformModel

noiseFractionUpperBound(OCT4knockout_betaUniformModel)

# If we were to use a P-value cutoff of 0.01, what would the FP rate be?
TP <- truePositiveFraction(OCT4knockout_betaUniformModel, pValueThreshold = 0.01)
FP <- falsePositiveFraction(OCT4knockout_betaUniformModel, pValueThreshold = 0.01)
round(100*FP/(TP+FP), digits = 1) # Only 2.8% FPs


# Perform meta-analysis using the TTRUST TF->target sets    
   
cas9OCT4_hsZygote_diffexDT[, betaUnifScore_FDR0.05 := betaUniformScore(koOCT4_pValue, OCT4knockout_betaUniformModel, FDR = 0.05)]

TTRUST_TF2targets_DT[, geneID := as.character(geneID)]

TTRUST_TF_pvalues <- cas9OCT4_hsZygote_diffexDT[ unique(TTRUST_TF2targets_DT[!is.na(geneID), .(TF,geneID)]), , on = 'geneID'][!is.na(koOCT4_pValue),.(pValueSet = list(koOCT4_pValue), geneSet = list(geneSymbol), scoreSum = sum(betaUnifScore_FDR0.05)), by = TF]

TTRUST_TF_pvalues[, betaUniformMixtureP := betaUniformPvalueSumTest(pValueSet[[1]], OCT4knockout_betaUniformModel), by = TF]
TTRUST_TF_pvalues[, fishersP := fishersPvalueSumTest(pValueSet[[1]]), by = TF]

TTRUST_TF_pvalues[p.adjust(betaUniformMixtureP, method = 'fdr') < 0.05] # No significant gene sets
TTRUST_TF_pvalues[p.adjust(fishersP, method = 'fdr') < 0.05] # Quite a few ... too many for so little signal?

TTRUST_TF_pvalues[scoreSum > 0]

adamsardar/metaDEGth documentation built on Sept. 28, 2022, 10:12 p.m.