Our task is to create an interface that scientists can load genome data in to, and mark regions of interest where the Machine Learning Algorithm will fit a model to the data. This model will take these regions where the DNA data is abnormal into consideration and dynamically generate new models as a scientists highlights more regoins The goal is that by using this model, doctors will be able to detect Cancer much earlier than they do now, since there is no curent way to analyze this DNA data other than by hand.
We plan to build a Web Application for scientists. This application will feature a way for scientists to upload DNA data into the system, as well as add labels to the data denoting where peaks should/should not appear. The data will be displayed in a similar look and feel to another popular Genome Browser, like JBrowse. The Machine Learning Aglorithm will draw a model to fit the data on top of the DNA data that will be updated as soon as a user uploads more labels about the data.
