GSK Data Management

A new data visualization interface.

Role Product Designer
Team MIT Media Lab
Company GSK
Time 10/2016-11/2016

Summary

GlaxoSmithKline (GSK) is the world’s sixth largest pharmaceutical company. In 2016, as a team member of the MIT Media Lab, I was invited by GlaxoSmithKline (GSK) to propose solutions to GSK’s data problems. The GSK Data Center of Excellence was implementing an initiative to bring the numerous disparate sources of data together into an integrated layer.

The visualizations presented herein were meant to build off an integrated layer – providing GSK employees a unified platform to search for information and to navigate between different entities, seeking out connections and relations, and absorbing information presented in an intuitive way.

The team of MIT, Harvard and Wellesley students visited GSK’s premises, interviewed various cross-departmental stakeholders from GSK, and designed a new interface.

Timeline

Understand Data

During the visit in GSK, we met and interviewed people from different departments, including Data Center of Excellence, Screening Profiling & Mechanistic Biology, Automation, Computational Modeling & Sciences, Molecular Design, Clinical Study Management, Oncology/Biology and IT. Areas touched on in our visit were mainly about Discovery and Development in the drug discovery process.

The current data solution from data center of excellence is to integrate all the data sources into one layer where users can access the data. Data sources from different stages of development will be used as the foundation. After data preprocessing such as standardization, data cleaning, joining, these sources are put into one integrated layer. Users can browse data in the layer as well as play with the data and create something new in the discovery zone.

To better understand data, we conducted research from platforms GSK scientists commonly used: ChemBL, OpenTargets, PubChem, ChemSpider and NCBI.

Current Challenges

Key Findings

1. Inability to integrate and review data analysis through all stages of development.

2. Difficulty in creating a collective learning environment.

3. Difficulty to parse and understand data.

4. Lack of flexibility between searching queries.

Target Audience

I categorized anyone with a role in the drug discovery process as “Players of the Game”. They are the target users for the design.

Design Principles

Search

Save more time and streamline the search process. The design will have a unified search across entities.

Connection

The page will present cross entity connections and allow the user to click on related entities and quickly navigate to them.

Brand Identity

The design uses a soft color palette that builds off GSK logo’s colors.

Design Overview

We identified the entities that are common across the 500+ database systems that GSK was using at that time. After several iterations, the common entities were defined, namely, compounds, targets, assays, clinical trials, animal trials, people and programs .

Workflow

Low-fidelity Prototypes

Design Details

Home Page

Search Summary

Profile

Workflow for Personas