Team: 5
School: Taos High
Area of Science: Medical
Interim: Interim Report
Supercomputing Challenge 2020-21
THS Team 5
Current methods of predicting and diagnosing cancer based on our genetic code are inefficient and inaccurate. We are building a tool to catalogue all current research into cancer in the genome and model known and unknown genetic causes and traits of cancer. This can be used to find new areas in the genome which may be studied in relation to cancer. It will also facilitate predicting and diagnosing cancer in individual patients based on their genetic code.
Our solution to this problem is to train a neural network on all current known genetic traits of cancer in humans, and then use it to study examples of genetic code for signs of cancer. Our deep learning system will be written in python using already existing libraries. This neural network will be trained on data about various genetic errors and related cancers sourced from the internet.
So far, most of our project has been in the research for the project. We have been exploring and learning as much as possible about the human genome and cancer, as well as how neural networks can be used in this area as well as in genetics and molecular biology in general. We have been working to lay the foundation of our neural net in TensorFlow.
We are expecting to meet our goals to a satisfactory level by the completion of the challenge.
Works Cited
Understanding Genetic Testing for Cancer. www.cancer.org/cancer/cancer-causes/genetics/understanding-genetic-testing-for-cancer.html.
Mostavi, Milad, et al. “Convolutional Neural Network Models for Cancer Type Prediction Based on Gene Expression.†The International Conference on Intelligent Biology and Medicine, 2019.
Louro, Javier, et al. “A Systematic Review and Quality Assessment of Individualized Breast Cancer Risk Prediction Models.†British Journal of Cancer, 2019.
Lodish H, Berk A, Zipursky SL, et al. Molecular Cell Biology. 4th edition. New York: W. H. Freeman; 2000. Chapter 9, Molecular Structure of Genes and Chromosomes. Available from: https://www.ncbi.nlm.nih.gov/books/NBK21700/
Larochelle, Hugo, et al. “Exploring Strategies for Training Deep Neural Networks.†Journal of Machine Learning Research, Edited by Leon Bottou, 2009.
Team Members:
Haven Hennelly
Sawyer Solfest
Max Meadowcroft
Sponsoring Teacher: Tracy Galligan