Synthetic and Null Data Generator (SaNDGen)

The Synthetic and Null Data Generator (SaNDGen) application is an R Shiny app developed by Parker Combs (NIH/NIEHS) and adapted from code by Deepak Mav, PhD (Sciome, LLC) with input from Scott Auerbach, PhD (NIH/NIEHS). Given an input file of a specific format (see ./examples/ subdirectory), the app can:

• Generate some number N of synthetic null samples based off of the existing null samples only in the input file.
• Generate some number N of synthetic null samples based off of the existing null samples only in the input file and randomly sample and categorize them into arbitrary dosage levels of size M_i ... M_j where M_i is the first defined dosage level and M_j is the last defined dosage level.
• Generate some number N of synthetic null samples for each dosage level in the input file and randomly sample and categorize them into arbitrary dosage levels of size M_i ... M_j where M_i is the first defined dosage level and M_j is the last defined dosage level.

Usage

1. First, click the "Browse..." button to open the file browser. Select one or more input files to upload to the app.


2. Next, select the output mode.


3. For the Generate synthetic null control samples mode, you can set the number of synthetic samples option to determine the total number of original + synthesized samples to be output.


4. For the Generate synthetic null control samples & create expression data files mode, you can set the following options:

• number of synthetic samples - The total number of original + synthesized samples to be output.
• number of expression data files - The number of output files to be created. Each output file will have a different sampling of the total samples specified in the number of synthetic samples.
• dose input mode - How the dosage levels for each sample should be generated.

5. For the Automatic generation dose input mode, you can set the following options:

• Number of treatment groups (including D0) - The total number of dosage groups, including the null dosage group (dose = 0).
• Number of replicates = D0 - The total number of samples per output file where dose = 0.
• Number of replicates ≠ D0 - The total number of samples per output file where dose ≠ 0 (this is applied to each nonzero dose level).
• Top dose - The maximum dose level to include in each output file.
• Dose spacing - The spacing between each dose level.

6. For the Manual input dose input mode, you can click the Add dose level button to specify a new dose level as well as the number of repetitions for each level.


8. For the Supplement existing data mode, you can configure the options similarly to the previous modes to generate new samples for each dose level given one or more input data file(s).


9. Press the GO! button to start the sample generation process.

10. Once the sample and file generation is complete, a new button will appear that allows you to download a .zip archive of the files generated.