Abstract:

Hierarchical clustering algorithms, such as Pearson’s correlation, Euclidean distance, Euclidean distance harmonic, Spearman rank correlation, Kendall’s tau, and City-block distance, were used to find the best way to establish theoretical MAPK/Erk signaling pathway on the basis of breast cancer line MCF-7 gene expressions. The algorithm constructs a hierarchy from top to bottom on the basis of a self-organizing tree. It dynamically finds the number of clusters at each level. It was found that only Euclidean distance harmonic is fit for the analysis of the cascade composed from a RAF1 (c-Raf), a MKNK1, a MAPKK (MEK1/2) to MAPK (Erk) in breast cancer line MCF-7. The result is consistent with the biological experimental MAP/Erk signaling pathway, and the theoretical MAPK/Erk signaling pathway on breast cancer line MCF-7 is set up.

Key words:

hierarchical clustering;signaling pathway; Euclidean distance harmonic