NETMAGE: humaN-disEase phenoType MAp GEnerator for the Visualization of PheWAS Results

NETMAGE allows for the automated generation of interactive disease-disease network visualizations from input PheWAS summary data. Source code is available at https://github.com/dokyoonkimlab/netmage, and example input data can be found at https://github.com/dokyoonkimlab/netmage/tree/main/example_data. The link to your DDN will be available for 24 hours, after which we delete all files. Please download the “.netmage” file from your resulting network if you want to recreate your DDN later – clicking the “Upload netmage file in place of summary files” button and using your .netmage file will let you generate your DDN again without having to re-upload your raw data. You can also always click the "back" button on your browser and change your input parameters to recreate your DDN without re-uploading the files. Feel free to check out our video tutorial, which gives a walk-through of how to use NETMAGE and how to interact with your DDN. For any issues with the website, please reach out to Manu.Shivakumar@pennmedicine.upenn.edu and for any further questions about the software, feel free to reach out to Dokyoon.Kim@pennmedicine.upenn.edu.

The “Upload PheWAS summary statistics” section can be used to input your own PheWAS data. Input file(s) must be text files (.csv, .tsv etc.), and can be uploaded either by dragging and dropping or by clicking and navigating to your source directory.

Upload PheWAS summary statistics

The “Parameters” section lets you specify the title and introduction of your DDN visualization, as well as your desired maximum p-value and minimum MAF thresholds. These latter two options allow you to vary the significance of variants being considered for data visualization, as well as ensure that your network includes primarily common variants. Furthermore, the “Node attribute file” option within this section lets you upload a tab-separated file of disease-to-disease category mapping and is used to introduce color by category into your DDN. The columns of this category must be “phenotype” and “category” respectively, and the phenotypes included in the “phenotype” column should match with the phenotypes specified in your input data for coloring to work. Additionally, the 'LD file' option can also be used to upload an LD file that will prune variants in your data. The LD file for your data can be generated within PLINK using the “show-tags” option. Finally, the “Marginalize edges” checkbox can be used to change the definition of edges in the network. If left unchecked, the weights of edges in the DDN are defined as the number of shared variants between the two phenotypes. Otherwise, edge weights correspond to marginalized fractions of variants – i.e., the number of variants that constitute the edge are divided by the union of the individual sets of variants for both phenotypes.

Parameters

The “Summary File Header” section lets you tell NETMAGE how your input data are formatted, asking for you to specify the column names for your SNP, phenotype, minor allele frequency, and p-value variables, as well as the delimiter used in your input data. Without including the appropriate information in this section, NETMAGE will not be able to properly read your input!

Summary File Header
* If this header is not found then the filename is used as phenotype name.
* Use \s for space and \t for tab.

The “Network Layout” section lets you specify the organization of your resulting DDN.

Network Layout

The “Additional Parameters” section includes extra options to manipulate the aesthetics of your resulting DDN as desired.

Additional Parameters
Legend
Features
Drawing Properties
Graph Properties
Mouse Properties