The Center for Spatial and Functional Genomics is a collaborative group at the Children’s Hospital of Philadelphia (CHOP) consisting primarily of Andrew Wells’ lab and Struan Grant’s lab. The center uses next generation sequencing technology to study the 3D configurations of genes and their interactions. Through bio-informatic analysis, the group is able to use these genetic libraries to uncover novel disease related mutations with the long term goal of understanding and treating auto-immune disorders and other conditions.
The kinds of assays needed to do this kind of research (ATAC-Seq, RNA-Seq, Hi-C, etc) create a large output of data that becomes very difficult to sort through. While the group has several very talented bio informaticists, it still leaves a lot of data that is under explored and under utilized. Bio-informatician Chun Su created two different data exploration web applications to enable researchers not trained in bioinformatics to be able to use this data to inform their studies and create figures for their research papers. Over a year after their creation however, the applications were not being used by the staff not trained in bioinformatics as they had intended to be. After Chun left CHOP, the project was handed over to Zachary Thompson, who asked me to help with the redesign.
In order to understand what needed to be changed about the app I created a flow chart of the existing processes and I did a heuristic design analysis of the existing interface.
.png)
Additionally, I observed 2 of my non-bioinformatics coworkers using the app. They were getting stuck at the LDLink API Token stage and one of their searches failed because they had the wrong input at the second search type selection input.
In revising this app, it was also important to remember the organization's higher level goal of eventual turning this internal app into a web resource for researchers around the world. Any ambiguous terminology needed to be clearly explained for potential students and researchers less familiar with genomic research.
Many of my recommendations could be achieved simply with the addition of some spacing, consistent typography, and some explainers. The problem that inspired the most design possibilities has problem #1 from my SNP menu analysis. This easily missed radio input was essential to executing a successful search but it was buried all the way at the bottom. During my observations, one of my users missed this button and had to try the search again. It was both very important but also completely redundant, since users make the choice of SNP vs GENE long before reaching the end of the crammed menu.
The developer I worked with on the design for the app was primarily a bioinformaticist with a limited coding skill set in R-studio. While he was enthusiastic about the designs I’d proposed, they didn’t line up clearly with his technical abilities. For the final version as of the last time we worked together, we landed on a compromise design that worked with what he was able to create before ultimately leaving the lab. Presented below is the final version that was passed along to the person taking over the project.
As someone who worked in the center on the creation of many genetic libraries, I was often curious about how I could see the fruit of my labor. Making this data accessible to the researchers who work in the center and ideally beyond is the ultimate goal in helping progress of congenital disease research. The product is out of my hands now and hopefully with a developer with greater R-studio capacity and time availability to apply the usability suggestions I prepared.
