Hello,
I am very new to the field of analysing gene expression. I actually have an excel table with the mRNA expression of my genes and i want to compare the expression for example let's say a gene B and a gene A.
So i know that the fold change is the value of B divided by the value of A (FC=B/A). i saw some tutorials but some people do the following formula after calculating B/A : logFC= Log(B/A) and then they do Log2FC= 2^-(LogFC).
is this right and this is enough to get the Log2 values? and if it is not how can i calculate the Log2 value? and a final question can the log2 value be negative?
Thank you so much!
First you will need to normalise your transcript data. See https://bmcbioinformatics.biomedcentral.com/articles/10.1186/s12859-015-0778-7
Then it is necessary to perform a log2 transformation to the normalised values, because transcript data is always very skewed.
You can use an empirical Bayes (eBayes) model to compare treatment groups (e.g. with R package limma).The eBayes model computes a moderated t-statistic test for each individual contrast (of your treatments) to test if they are equal to zero (this is the hypothesis test for difference). As mentioned above.
This produces an F-statistic, which is an overall test computed from the set of t-statistics for each transcript. This is analogous to the relationship between t-tests and F-statistics in a conventional analysis of variance, except that the residual mean squares and residual degrees of freedom have been moderated between transcripts. Data with a probability (P) value of < 0.05 and a log2-fold change (FC) < -1 or > 1 are 'important'. But, all comparisons with p-values need to be adjusted using something like the Benjamini-Hochberg method with a false discovery rate (FDR) set to P ≤ 0.05.
Good luck!
- Badges
- Science topic
- Similar topics
- Engineering
- Materials Engineering
- Powders