DREAM Data Analysis Tool

Last year, DREAM and the National Cancer Institute's Clinical Proteomic Tumor Analysis Consortium presented a challenge to the open science bioinformatics community. This challenge, the NCI-CPTAC DREAM Proteogenomics Challenge, came with three problems. Using real data from ovarian cancer and breast cancer patients collected by NCI-CPTAC, the competition awarded teams for (1) accurately imputing missing protein abundance values, (2) predicting protein abundances given mRNA and DNA data, and (3) prediction phosphoprotein abundances given protein abundances as well as mRNA and DNA data.

The DREAM Data Analysis Tool, designed by Team 30, will provide solutions to these problems with functionality that can be accessed entirely through a graphical user interface. The tool will use known algorithms as well as algorithms developed by Team 30 to solve the problems. The Data Analysis Tool will be designed to be easy-to-use to any biological researcher, eliminating the need for a competent programmer to be on their research team.

Data will be uploaded from the client’s computer to the Data Analysis Tool Server, and will be displayed on a spreadsheet for the user. From there, the user will be able to select visualizations and analyses for their data from a drop-down menu without learning how to code. The user can also choose from a selection of machine learning algorithms to analyze their data. Although the focus is on the three subproblems presented by the NCI-CPTAC DREAM Proteogenomics Challenge, the eventual goal is to allow the user to use machine learning to analyze any data set.

Project-related Resources:

• DREAM Challenge NCI-CPTAC Proteogenomics Challenge Page
• Missing value estimation methods for DNA microarrays. Troyanskaya et al.
• Missing value estimation for DNA microarray gene expression data: local least squares imputation. Kim et al.
• Stanford University’s Coursera course on Machine Learning
• Missing data imputation using statistical and machine learning methods in a real breast cancer problem, Elsevier

Useful Links:

• Data Analysis Tool (Coming Soon)
• Video tutorial using the Data Analysis Tool