Our Space Experiment Visualization Tool offers a comprehensive platform for users to explore and understand the intricate world of biological experiments conducted in space, specifically focusing on datasets OSD-379 and OSD-665 from NASA's Open Science Data Repository (OSDR). The scalable tool provides a variety of valuable features, including dynamic visualizations of experimental designs, results, and timelines, as well as contextual information to enhance user comprehension. Users can access compelling visual representations of the experiments, illustrating critical elements such as treatment groups, timelines, and the number of subjects involved, allowing researchers to quickly grasp the structure and outcomes of the experiments.

Designed with scalability in mind, our tool seamlessly integrates additional datasets from the OSDR. As new datasets are released, they can be incorporated into the visualization tool without major modifications, ensuring that it remains relevant for ongoing and future research. The platform is enriched with interactive capabilities, such as filters and tooltips, enabling users to delve deeper into specific datasets and enhancing user engagement. To further improve user experience, we have integrated a chatbot that assists users in navigating the tool and answering common queries related to the experiments and visualizations. This feature provides contextual information, guiding users through the functionalities and enhancing overall usability.

Here is an overview of key features of our user interface:

1. Landing Page

The user begins their journey by navigating to the Space Experiment Visualization Tool website. Upon arrival, they are presented with a user-friendly interface that invites interaction. Users can easily paste the URL of their desired dataset from the NASA Open Science Data Repository (OSDR) into the provided input field (see figure 1).

upload_image.03:13:12.591238

Fig 1: Homepage and Link Submission

2. Interactive Dashboard

Upon successful data processing, users are seamlessly redirected to the home page of the application. At the top, the project title is prominently displayed, providing immediate context for the user. The layout is structured with two rectangular panels positioned side by side (see figure 2).

Experiment Information Panel (Left): This section features a navigation bar with intuitive buttons for accessing various functionalities and detailed explanations of the respective experiments.
Visualizer Panel (Right): This panel presents a detailed visualization of the biological experiments associated with the selected dataset. The timeline includes interactive nodes that act as filters, allowing users to explore specific aspects of the experiments. By selecting these nodes, users can delve deeper into the data, accessing visually compelling representations for a more comprehensive understanding of the experiment results.
Chatbot: A chatbot is available as a pop-up in the interface, providing real-time assistance to users as they navigate through the application. The chatbot offers instant feedback, enhancing the overall user experience by answering questions and guiding users through the tool

upload_image.00:06:01.014838 Fig 2: Interactive Dashboard Overview

3. Interactive Diagram Features:

Tree-Like Structure: A clear branching layout, where each branch represents different experimental conditions (factors) applied to various groups. This gives a visual representation of the experimental flow and factor distribution (figure 3).
Unique Factor Icons: Each factor in the experiment is visually represented by a distinct icon that is dynamically generated based on the experiment data types, allowing for quick identification and differentiation between variables (independent variables and groups).
Zoom & Scroll Functionality: Users can zoom in to examine specific parts of the experiment tree closely or scroll across the diagram to explore the structure at a broader level (figure 4)
Factor Value Sorting and Tree Re-structuring: After scrolling to the bottom of the visual, there are drop downs that allow you to filter by factor value. For instance, filtering by Space Flight and On ISS restructures the tree to show only sample groups with those factors (figure 5)
Interactive Result Pages: Clicking on the end of each factor chain expands the result page, offering detailed visualizations of the outcomes and population data for the sample groups affected by the applied factors.
Quick Group Comparison: The linked tree shows how different sample groups are structured based on their assigned variables, helping users understand the relationships and differences between groups at a glance.
Reset & Save Options: These buttons allow users to reset the visualization to its default view or save their current arrangement for later review, aiding in usability and data management.

upload_image.00:08:13.443814 Fig 3: Study Visualizer Full Tree

upload_image.00:08:00.404464 Fig 4: Close up of a branch

upload_image.03:11:24.332672

Fig 5: Example of a filtered tree based on Space Flight and On ISS factor values

4. Results Pop-Up

At the far right ending of any branch in the visualizer tree (shown in figure 6), users can navigate to a pop-up that displays the results for a specific sample and its associated factors. This pop-up contains various visualizations that allow scientists to quickly and effectively interpret the data (See figure 7).

Key visuals include:

Sample Composition: Offers a breakdown of the sample’s key components.
Correlation Heat Map: Helps identify relationships and correlations between different parameters within the sample.
Mean Parameter Values: Provides an overview of the average values for relevant experimental factors.
Distributions: Displays the distribution of parameters, enabling a more detailed understanding of variations within the data.

These interactive visuals help scientists easily grasp complex data sets and extract actionable insights.

upload_image.00:05:25.831567 Fig 6: Navigating to results page from the tree visualization

upload_image.00:04:01.699966 Fig 7: Experimental Results Pop-Up Visualizer

4. Experimental Information Panel

The left panel is an interactive set of buttons that each given different insights on the experiment. For instance, buttons like the "NASA Center" or "Organism" dynamically link to a Wikipedia page, allowing users to quickly learn more about the background information of the study (see figures 8 and 9).

upload_image.00:13:00.753748 Fig 8: Left panel overview on dashboard

Additionally, buttons within this panel like "Show PI Contact Information" and "Sample Characteristics" have a pop-up which displays the information about the study. An example is shown below:

upload_image.23:47:58.068221 Fig 9: Example of button pop-up for sample characteristics

5. Interstellar Automated Visualizer Bot

If other questions arise regarding the experiment when viewing the visualizations, the chatbot button on the right of the screen can be clicked, and a chatbot text bar will appear. Questions regarding the experiment, for instance "What are the factors in this experiment?" can be entered, and promptly generated and answered, allowing for an interactive experience. Figures 10 and 11 show examples of questions that can be asked.

upload_image.03:37:22.242286

Fig 10 & 11: Example questions for IAV Bot

6. Technologies Used

Our tool uses a combination of advanced technologies including:

Python
React
JavaScript
CSS
Chatbot Kit
Axios
Playwright
Plotly
Dash.
Langchain
OpenAI

2024 NASA Space Apps Challenge

Space Meatballs