The experiment main article.  What we learned from the experiment and what we discovered. If the light bulb is on it means we believe it is an important experiment. The expression data. Probes in rows and Samples in columns. The expression data profile. This file contains all the necessary information about Samples.  A density plot for the expression data. The training set for top 100 probes. First, we randomly split the expression data to training set (80%) and test set (20%) and then use Mahalanobis distance to choose the top 100 probes. Finally, we slice both training set and test set using the top 100 probes. The test set for top 100 probes. It is important to split the expression data to training set and test set before applying a statistical test on the training set to choose statistically significant probes.  A link to Kaggle.com. We occasionally upload the data to Kaggle for public discussion.