Hello, my name is Graciany Lebron, and I am originally from Bayamon, Puerto Rico but I study Computer Engineering at the University of Puerto Rico at Mayaguez. Here are some facts about me if you are interested:
This summer I am working with Dr. Vetria Byrd and Dr. Mark Daniel Ward on a project/pilot study that will give insight to lupus medical professionals and people in general about SLE using big data and visual analytics. This project as a whole will consist of a web based framework that will ultimately be a good resource to doctors since all of the data related to lupus will be gathered in one place in a textual and/or visual form. As a result, this will likely cause a decrease in the time between diagnosis and treatment. However, to achieve this, we have to face the challenges brought by the heterogeneous nature, volume, and uncertainty of Lupus data. This Pilot study will be the first step towards the development of a flexible, but robust framework that will allow for the integration and visualization of heterogeneous data.
During this week there were many introductory talks about the program, graduate school, and research as well. We were told by the program directors and coordinators what to expect through out the summer and had a tour around the university. Additionaly, I met both of my mentors and talked about the anticipations they had for the research this summer. Shortly after that, I started working, by taking the CITI program training to get familiarized with the ethical aspects of research, and by reading introductory papers about Data Visualization which is the area of expertise of the PI that I'm working with.
Dr. Byrd and I, during this week defined the project from a list of topics that Dr. Byrd had thought out. The project that was picked combined all of the areas that my research mentor wanted to work on which were big data, uncertainty visualization, and heterogenous data, and the type of data that I wanted to work with, which was biological. Therefore, it was decided to visualize lupus data. Before defining the milestones, I had to read a paper about software engineering recomended by Dr. Byrd that should help us define them and structure the systemic literature review of this project. Meanwhile, I searched and pulled articles that were referenced in a survey paper about heterogenous data as a starting point for the literature review. On the other hand, Dr. Byrd also gave us some exercises in excel about working with pivot charts, and encouraged us to take a tour to the envision center with Dr. George Takahashi to see how data visualization and virtual reality can work together.
Throughout the course of this week, Dr. Byrd and I established the milestones of the research. Thus, the scope of the topics, the research questions, the search process, and the visualization framework for the research project were defined. For the purpose of having a better idea of what things could be done during the span of this summer. The first task was to do a systematic literature review of SLE, using the research questions to filter which articles should be chosen above others and the search process to find papers in journals that emphasize on lupus. At the same an introduction to html was given by Dr. Byrd and we started working on our personal webpages.
At this time Dr. Byrd had to go and give class in Colorado for the week. She advised me to go and talk to Dr. Ward about the status of the project to keep him updated and to see what he thought. When I visted Dr. Ward, he suggested that I focused on making a robust data visualization framework. Later on during the day, Dr. Byrd and I talked on the phone to discuss my meeting with Dr. Ward and to also give her update on the project. After talking, we agreed to focus on the framework as well. For now most of the systemic literature review has been paused but will continue to look for more information on the side while I work on the framework itself. Here I explored which data visualization tool to use for the project, I tried using Cytoscape, Cytoscape.js, Tableau, Gephi, and D3.js. But finally opted for D3.js since it is an open source software that had the most documentation compared to the other tools. Overall, this was a rough week for me since there im new to D3.js and had trouble setting up the environment for the framework.
This week after some delay I was able to create the first prototype of the lupus framework where you can see the different information about the disease such as: the associated genes, the environemental risks, the symptoms the treatments, the myths, the diseases it imitates, and the disparities, more topics will be added later on. For now in this pilot study, since we are working with lupus published literature all of the files are in pdf format and the data extracted from these files are of text and numeric data types. because of the private nature of lupus databases and the scope of the project is too broad. Although, in the future when we acquire data from various database sources we will be working with other types of data such as images, xml files, pdf files, and much more. By doing this we will hopefully create a visualization framework that is able to deal with heterogeneous data. Aside from that, I was able to start developing the associated genes view, and write a rough draft of the abstract that will be due the friday of next week.
In the main view of the framework, the placeholder text that was in each of the modules in this view were replaced with lupus data related to each of the topics that we chose to focus on as a starting point. In addition, the symptoms subview was created and the data visualization associated to this view were designed. While all of this was being done, the sources of each view were cited since all of this information was acquired from lupus published work. On the other hand, the treatment view was started and the basic visualizations such as the bar chart and pie chart were explored in the associated genes just to reaffirm that they do not represent the nature of the data accurately.
The focus of this week was to create the treatment sub-view of the framework, which was the last sub-view that I was able to create due to the time constraint of the summer program. After that, I was able to finish the first prototype of the framework. Because of that, I was able to continue to edit my final presentation and have a draft for the poster by the end of this week to have everything set and ready for the activities of the final week of the program.
This is the last week of the program, here I had to finish the final presentation and poster since they were due and they had to be presented this week. After many drafts, I was able to complete all of the tasks at hand and I was able to show everyone what I've being doing through out the whole summer. Overall in this summer research experience, I have learned many things that were out of my discipline, did things that were out of my comfort zone, and met new people. For all these reasons I thank Dr. Byrd and Dr. Ward for giving me this great opportunity to work with them and to be part of the SROP program this year.
The goals of this pilot study during the scope of this summer can be summarized as follows: conduct a preliminary literature review on SLE, search visualization tools that integrate and visualize heterogeneous data, and create a prototype of the framework. All these goals were designed to take the first step towards the main goal of the whole project which is to explore new methods to integrate and visualize heterogeneous lupus data, and reduce the time between diagnosis and treatment by giving insight to lupus medical professional using visual analytics.
Last updated: 07/29/2017