Understanding the heterogeneous nature of cancer tissue is a very important problem in cancer research. It can give insights into the cause of disease, its progression and explain induced drug resistance. There are two models that are used to explain heterogeneity, Cancer Stem Cells and Clonal Evolution. This thesis aims to address this challenge by developing an algorithm to determine the ratio of different components of a heterogeneous cancer tissue. This algorithm is robust and does not depend on the heterogeneity model. The proposed algorithm uses response vector, which is a vector of observable response of cell lines. A database of the response of individual cell lines is developed by collecting cell-by-cell response measurements. A heterogeneous cancer tissue is modeled as being a mixture of these cell lines. Avoiding the high cost cell-by-cell analysis, the collective response of the heterogeneous cancer tissue is observed. The algorithm uses Bayesian inference to estimate the probability distribution of the number of cells of individual cell lines based on the response of individual cell lines and the observed collective response. The results of the algorithm are validated using synthetic data and real-world data collected from cell lines, which are mixed in a ratio known a priori.
