Comparison with k-means when k=2

Below is a GO enrichment comparison between k-means with k=2 and the ten significant model profiles out of 50 with c=2. The silhouette method implemented in Matlab determined that k=2 is the optimal value for the number of clusters for k-means. However running k-means with k=2 did not lead to any significant GO category that was more enriched than with the profiles method.