I need an advice. I have a dataset consisting of 108 observations (27 subjects * 4 time points) and ~10000 features. The data represents intensity values (comes from continuous domain). When I run unsupervised methods such as Hierarchical Cluster Analysis or Principal Component Analysis, I observe that I can reveal clusters consisting of data points belonging to the same subject. I assume that there are certain features that lead to clustering.
I would like to ask how could I retrieve the features that lead to separation between subjects with respect to the fact that there are only 4 data points per subject and 27 groups/subjects?
I want to point out that I don't only speak about unique features but about features that are possibly shared among several subjects.