Developing a clustering algorithm to analyse flow cytometric data

Abstract

The analysis of flow cytometry data requires identification of cell clusters. This clustering is done manually by applying gating techniques and by analyzing the plot scatter across each dimension. Since all of this is done manually it becomes really hard to find some rare event or to identify the cluster. The aim of this project is to develop a clustering algorithm that would automate the process of cluster formation on large flow cytometric data.Flow cytometry (FCM) is a technique used to detect and measure physical and chemical characteristics of a population of cells or particles.(Wikipedia)The cells are placed in a column filled with saline(NaCl+H2O).The cells exits the column one at a time. Saline helps in this process through hydro focalization’s soon as the cell exit the column a laser beam hits them and the cell scatters the laser beam in accordance to its size and complexity.