.. (לתיקייה המכילה) | ||
In question 2, should the clustering be done on rows or columns? | |
As this is one of the questions you are asked (2e), we cannot tell you explicitly what the answer is, but we can further explain: You are asked to do clustering on the *genes*, in a way that will differentiate the samples into their 2 conditions (samples 1-5 is one condition , and samples 6-10 is the second condition). As we learned in class, in the clustering process we find vectors of values that are similar to one another. To understand what it means to differentiate between the conditions, take a look at slide 9 in lecture 6 - in that slide you can see a set of genes that express differently between the two conditions (ES cells and Fibroblasts in that case). There may be another set of genes that won't show such a different expression pattern between the 2 conditions, for example, they would be highly expressed in all the samples. Another hint to this question can be found in the final question: "What is the optimal number of clusters to find *groups of genes* that differentiate between conditions". |