Hello!
I'm trying to reproduce results from an article on RNAseq clustering. Authors use NMF for unsupervised consensus gene clustering. And than "A preferred cluster result was selected by considering profiles of cophenetic score and average silhouette width (Rousseeuw 1987) of the consensus membership matrix, for clustering solutions having between 3 and 15 clusters (Fig. S5.1b). Silhouette results were generated from the NMF consensus membership matrix using the R ‘cluster’ package v1.14.1." Unfortunately, I don't understand how do they get the result from NMF clustering using cluster package. Could smn, please, kindly explain it to me? I am new to RNAseq. I will be very thankful for any help on this.
My workflow: I downloaded normalized TCGA file, chose 1500 genes with the help of Median Absolute Deviation (MAD) rna_seq_data_mad_genes_sel = FSbyMAD(rna_seq_data_modif, cut.type = "topk", 1500)
than I performed NMF clusterisation res <- nmf(rna_seq_data_mad_genes_sel, r=50, 'brunet', nrun=2, .options='v4')
Check out this blog: Unsupervised clustering on gene expression data.