What percentage change in methylation of cpg sites is biologically meaningful
I was reading a paper and was surprised to see that many of the genes listed as significantly demethylated had ~1% differences in cpg island methylation.
can somebody suggest a paper that can explain what seems to be a small effect size is biologically meaningful
• 2,407 views
•
link
written
by
confusedbiologist •
0 answers
No answers yet.
Log in to answer this question.
More posts like this
-
PAML/codeml Poxvirus dN/dS
written by Adrian Pelin
-
How to annotate DMVs in methylation data (via RHmm) ?
written by Bosberg •
-
Working with different batches: when to apply normalization
written by mincej20 •
-
Probabilistic modeling of CpG ß-values
written by lotus28
-
Normalization of raw Illumina reads.
written by robert.murphy •
-
HISAT2/StringTie/Ballgown for Novel Transcript Isoform Differential Expression
written by ra711 •
-
Using StringTie with CAGE-seq data
written by nhaus •
-
Methylation at probe level and gene expression correlation
written by Will •
-
EdgeR analysis p-value plateau formation
written by Laura •
-
Beta-values, M-values and thresholds on effect size
written by gbayon
I may not be fully understanding the table, but I am not sure how those can be significant. It's very rare that a change of less than 5% and especially 1% can be significant and yet there they are.
Most of the expression differences are less than 20%, which is also very dubious for microarray-based experiments.
Such a paper doesn't exist, because such piddly changes are biologically meaningless. You'll find a surprisingly large number of "high impact" papers making a big deal out of meaningless gene or promoter level changes, since it's easy to get a nice p-value for them.