ref_compare | R Documentation |
ref_compare
computes log2-fold changes (logFCs) between given sample groups
and logFCs between matching expression profiles of a developmental reference.
By comparing the logFCs of the samples and those of the reference, we can
quantify expression changes caused by developmental differences between sample groups
rather than by the variable of interest.
ref_compare(X, ref, group, ae_obj = NULL, ae_values = NULL)
X |
sample gene expression matrix (should be comparable to reference, usually log(TPM+1)). |
ref |
a |
group |
a factor defining sample groups to compare (e.g. wild-type & mutant), the first factor level is used as control. |
ae_obj |
an |
ae_values |
age estimate values (used if ae_obj is NULL). |
We use TPM instead of raw counts to compute logFCs because TPM are more comparable across datasets, and because reference data is interpolated from and to log(TPM+1). Because of this, our logFCs may differ from those computed by differential expression (DE) analysis tools that use raw counts (e.g. DESeq2, edgeR)
an rcmp object.
requireNamespace('wormRef', quietly = TRUE)
# get sample gene expression data
X <- wormRef::Cel_larval$g[,1:9]
# get reference
ref <- prepare_refdata(ref = 'Cel_larval', datapkg = 'wormRef' , n.inter = 200)
# define groups
fac <- factor(c('a', 'a', 'b',
'a', 'b', 'b',
'c', 'c', 'c'))
# estimate sample age
ae_X <- ae(X, ref)
# compare group diff. expr. with matching reference
rc <- ref_compare(X, ref, fac, ae_X)
print(rc)
# get sample and reference (ie. development) logFCs between groups
lfc_a_vs_b <- get_logFC(rc)
lfc_a_vs_c <- get_logFC(rc, l = 'c')
lfc_b_vs_c <- get_logFC(rc, l0='b', l = 'c')
# plot sample vs. reference logFCs
par(mfrow = c(2,2))
plot(ae_X, groups = fac)
plot(lfc_a_vs_b, main = 'a vs. b logFC')
plot(lfc_a_vs_c, main = 'a vs. c logFC')
plot(lfc_b_vs_c, main = 'b vs. c logFC')
Add the following code to your website.
For more information on customizing the embed code, read Embedding Snippets.