Estimate the log fold changes based on a contrast matrix, requires the LFC package.
Usage
LFC(
data,
name.prefix = mode,
contrasts,
slot = "count",
LFC.fun = lfc::PsiLFC,
mode = "total",
normalization = NULL,
compute.M = TRUE,
genes = NULL,
verbose = FALSE,
...
)Arguments
- data
the grandR object
- name.prefix
the prefix for the new analysis name; a dot and the column names of the contrast matrix are appended; can be NULL (then only the contrast matrix names are used)
- contrasts
contrast matrix that defines all pairwise comparisons, generated using GetContrasts
- slot
the slot of the grandR object to take the data from; for PsiLFC, this really should be "count"!
- LFC.fun
function to compute log fold changes (default: PsiLFC, other viable option: NormLFC)
- mode
compute LFCs for "total", "new", or "old" RNA
- normalization
normalize on "total", "new", or "old" (see details)
- compute.M
also compute the mean expression (in log10 space)
- genes
restrict analysis to these genes; NULL means all genes
- verbose
print status messages?
- ...
further arguments forwarded to LFC.fun
Value
a new grandR object including a new analysis table. The columns of the new analysis table are
- "LFC"
the log2 fold change
Details
Both PsiLFC and NormLFC) by default perform normalization by subtracting the median log2 fold change from all log2 fold changes. When computing LFCs of new RNA, it might be sensible to normalize w.r.t. to total RNA, i.e. subtract the median log2 fold change of total RNA from all the log2 fold change of new RNA. This can be accomplished by setting mode to "new", and normalization to "total"!
Normalization can also be a mode.slot! Importantly, do not specify a slot containing normalized values, but specify a slot of unnormalized values (which are used to compute the size factors for normalization!) Can also be a numeric vector of size factors with the same length as the data as columns. Then each value is divided by the corresponding size factor entry.
Examples
sars <- ReadGRAND(system.file("extdata", "sars.tsv.gz", package = "grandR"),
design=c(Design$Condition,Design$dur.4sU,Design$Replicate))
#> Warning: Duplicate gene symbols (n=1, e.g. MATR3) present, making unique!
sars <- subset(sars,Coldata(sars,Design$dur.4sU)==2)
sars<-LFC(sars,mode="total",contrasts=GetContrasts(sars,contrast=c("Condition","Mock")))
sars<-LFC(sars,mode="new",normalization="total",
contrasts=GetContrasts(sars,contrast=c("Condition","Mock")))
head(GetAnalysisTable(sars))
#> Gene Symbol Length Type total.SARS vs Mock.LFC
#> UHMK1 ENSG00000152332 UHMK1 8478 Cellular -0.4395433
#> ATF3 ENSG00000162772 ATF3 2103 Cellular 2.4486712
#> PABPC4 ENSG00000090621 PABPC4 3592 Cellular -0.1344397
#> ROR1 ENSG00000185483 ROR1 5832 Cellular -0.1936823
#> ZC3H11A ENSG00000058673 ZC3H11A 11825 Cellular 0.7000287
#> ZBED6 ENSG00000257315 ZBED6 12481 Cellular 0.6889175
#> total.SARS vs Mock.M new.SARS vs Mock.LFC new.SARS vs Mock.M
#> UHMK1 4511.7770 1.1881622 1118.9972
#> ATF3 397.5641 2.8216633 341.9156
#> PABPC4 6301.4993 1.1819850 2062.9733
#> ROR1 2643.8914 0.6859184 1073.7033
#> ZC3H11A 3028.9232 1.1289929 1538.0788
#> ZBED6 2976.2018 1.1362473 1534.1663