Sometimes, there will be small technical difference between experiments, e.g. how active genes are in a sample. In these case, we need to normalize the data to compensate this technical difference.
The name comes from the fact that, the normalized datasets have identical quantiles.
How does it work
From the highest value from all samples, get the mean value and extend to new plot
From the 2nd highest value from all samples, do the same
Repeat the previous steps
After quantile normalization, the order get preserved, even though the value for each sample are the same.