Expression Heatmap Construction

the heatmaps has been constructed based on a subset of datasets restricted to high throughput sequencing (bulk rna seq) experiments normalization of expression values has been performed on a per dataset basis where samples were rank normalized and genes z score normalized for the combined datasets, an additional quality step was performed removing any dataset showing high batch effect upon pca analysis within mavatar discovery, you can choose between supervised or unsupervised clustering during unsupervised clustering, samples are clustered using euclidean distance and average linkage clustering whereas genes are clustered using biweight midcorrelation and average linkage clustering read more about how to interpret the heatmap under patient stratification docid\ hr9we2av3ea3krik6ge p