Chalmers Conferences, 9th European Conference on Mathematical and Theoretical Biology

Modeling Distribution of Reads from Leukemic Cells DNA through Mixture of Truncated Poisson Distributions
Nicolas Flores Castillo

Last modified: 2014-06-09

Abstract


About a year ago, Prof. Yuriy Fofanov and his team from the University of Houston sequenced DNA material obtained from the laboratory of Dr. Michael Andreeff at MD Anderson Cancer Center from different types of cells in patients with leukemia (normal stroma (NS), bone marrow (BM) and bone marrow stroma (BMS)). Then, he and his team considered the number of left-hand end reads Y that cover a random site in a nucleotide contained in a region of interest of the sequencing project. Unfortunately, the zero counts of Y are unknown making the modeling of Y more complicated and sometimes not even feasible. In this talk, I will present a modified version of the EM-algorithm to model the distribution of Y using a mixture of truncated Poisson distributions. Although this is still work in progress, we believe that knowing these distributions will help to characterize several properties of the genome in these cells; for example, identifying cells that have a small number of mutations versus the ones that have a large number of them. More specifically, account for the number of different types of mutations in a sample of leukemic cells.


Keywords


Truncated-Poisson-Distributions; EM-algorithm; DNA-reads