The “TGM pipeline” is a workflow combining tranSMART server, Galaxy Server and MINERVA platform to enable visually-aided exploration, analysis and interpretation of high-throughput translational medicine data. We describe it in the article submitted to the “Big Data” journal special issue on Visualization in Data Science: “Integration and visualization of translational medicine data for better understanding of human diseases” by Venkata Satagopam, Wei Gu, Serge Eifes, Piotr Gawron, Marek Ostaszewski, Stephan Gebel, Rudi Balling, Reinhard Schneider.
The tranSMART part of the pipeline allows for exploration of integrated translational medicine data and for expert-based selection of important sub-groups. These are pipelined directly to the associated instance of Galaxy Server.
No login is needed, the user will be automatcally logged in as “guest user”, allowed to export data to Galaxy.
In the tranSMART interface, user can navigate terms from the “Navigate Terms” panel on the left side of the interface and click the folder structure (or “data tree”) to explore the studies and variables (“concept”) for each subject in the study (Step 1 in Figure 1).
Then one can drag and drop each concept to the right panel into either “Subset1” or “Subset 2) to build customized cohort for comparison based on the filtering of the chosen concept (Step 2 in Figure 1).
After customized cohort is defined, then using the “Data Export” function in on the top of the interface, user can export the selected data set (Step 3 in Figure 1).
Once the export is finished, user can click the “Galaxy Export” to export the dataset to the Galaxy server. In this step a name should be given by the user as a identifier of the dataset in the Galaxy server (Step 4 in Figure 1).
Figure 1. The tranSMART interface and steps to follow.
The Galaxy part of the pipeline allows for establishing analytical workflows for data from the tranSMART server. Analytical results (list of molecules) are sent to the associated MINERVA instance, hosting the PD map.
login: firstname.lastname@example.org password: transmart
In the Galaxy interface, first click the “Shared Data -> Data Libraries” from the top (Step 1 in Figure 2). Then a list of data set will be shown. User then can click on the dataset that he/she just exported from tranSMART (Step 2 in Figure 2) and import them to the current history (Step 3 in Figure 2)
Then by clicking on the “Workflow” on the top of the interface and choose “tranSMART_PDMap -> Run” (Step 4 in Figure 2), one can choose the dataset to be analyzed using this workflow.
Figure 2. The Galaxy interface to choose dataset and workflow.
In the next step step you can choose the imported dataset as inputs for the workflow and run the workflow (Step 1 in Figure 3). After the analysis is finished, one can view the results by click the eye-shape icon (“View data”, Step 2 in Figure 3). There one can access the overlaid results on the PD-Map (Step 3 in Figure 3).
Figure 3. The Galaxy interface to view results and link to PD-Map.
The code explaining the construction of request to MINERVA-hosted PD map can be found here.
The MINERVA part of the pipeline allows for interpretation of analytical results from Galaxy Server in the context of hosted molecular interaction networks. Here, Galaxy results are projected on the PD map, and Parkinson’s disease-related mechanisms.
As shown above, you will be taken directly to the website from the Galaxy Server. You can also access the site directly using the link above and the following credentials:
login: galaxy password: galaxy.
For more information, please, contact:
Wei Gu (email@example.com)
Venkata Satagopam (firstname.lastname@example.org)