We have lost our melt calibration file for HRM analysis during software update and can't use precision melt analysis software for data analysis. We are using R programming based analysis for HRM. However, facing problem regarding clustering issue. Can't be confirmed about clustering is actual or not. How to decide optimal number of clusters for k-means and hierarchical algorithms, it seems sometimes confusing. Getting new calibration using kit is not possible currently.

If someone give the melt calibration file, I may try to check if the data can be generated from software or not and may compare with R analysis data.

Initial melt curve from CFX96 Real Time machine is also not smooth, shown in attached file. Please suggest what are the probable reasons and how can be solved this issue?