Challenge Team Abstract
Team Number: 003
School Name: Sandia Preparatory
Area of Science: Biogenetics
Project Title: Parallel Processing of Human Genomic Pediatric Leukemia Data Using Neural Networks and Genetic Programming
Purpose of our program:
There are 150+ subtypes of leukemia postulated to exist. The researchers are trying to identify them because they each have unique clinical courses, and respond in unique ways to different treatment regimes. This will save leukemia patients from enduring treatments that are less likely to be helpful, and some of which have terrible side effects. Our program will be a small step towards this larger goal.
We are going to program using either neural programming or genetic programming to create a grouping program to try to identify some of these subtypes.
The program will take input data consisting of raw numeric data of the levels of gene expression from cDNA microarrays, obtained from leukemia patient tissue samples, and compared with normal samples.
Our first input datasets will be too small to resolve more than a few subtypes, but will provide a good basis for testing our grouping methods.
Program development plan:
Our program will attempt to group the data, each group containing data from a different type of leukemia, according to which a specific pattern of genes was expressed in each tissue sample in the input data.
If our program is shown to be accurate with this small data set, then we will expand it to accomodate a larger data set potentially containing more types of leukemia, depending on what data sets are available by then.
This is new and ongoing research. There are already some known data sets from other labs that we can use to test our program. Hopefully there will be some data from the UNM research lab that we will be able to use before the end of this competition.
Our program will ideally be set up to run on parallel computers, from the outset.
Our output data will be numeric also. We will use the Vx-Insight graphics and analysis program developed by researchers at Sandia National Laboratories, who are working with the UNM group, to display the results of our program.
- Cheryl L. Willman, MD
- Susan R. Atlas