With Rosalind's Sample Re-use service, you can create a new analysis from the processed sample data you already unlocked for RNA-seq or smallRNA-seq
Step 1: Confirm your samples are eligible
Currently, this service is only available for RNA-seq and smallRNA-seq for paid annual subscription levels. Additionally, each sample you would like to reuse needs to have already been uploaded and processed in ROSALIND (samples with unprocessed FASTQ files do not qualify). There are 2 common scenarios for sample re-use requests:
- Scenario A: "I would like to re-use a subset of samples that were all originally uploaded in the same experiment" or "I would like to remove an outlier sample from the analysis"
- In this case, please go to the guide linked here and follow the instructions for Option 2
- Scenario B: "I want to re-use 5 samples that have already been processed together with 15 unprocessed FASTQ samples" or "I want to re-use samples that have all already been processed in a ROSALIND experiment"
- In this case, you would upload any unprocessed FASTQ samples first and unlock their analysis. Once all the samples required for reuse are already analyzed and unlocked, you will then be able to download the raw counts and attributes needed in order to manually combine them in a new experiment for your sample re-use. For further instruction, please continue to Step 2 in this guide:
Step 2: Obtaining the data
Go to the Download section of each individual experiment that contains the sample data you want to work with and open the "ROSALIND Ready for re-uploading as Processed Counts" folder, then download the Raw counts and the Attributes Files.
Step 3: Creating your new count file (2 options)
Depending on your comfort level you may choose to do this step yourself, otherwise please contact our Customer Support Team and we are happy to help!
Option 1: Contact Customer Support for assistance. Email the following to firstname.lastname@example.org and we will merge the data for you and then send it back once completed and ready for uploading
- All of the Raw Count (.TXT) downloads for each experiment from Step 2 above
- A list of the sample names you want included
Option 2: Create your own file, below are steps and tips for how to do this in Excel:
1. Open Excel and import each of the .TXT Raw Count files (downloaded in Step 2) into a workbook, the gene names should already be sorted in alphabetical order for you
2. Compare the data rows and make sure it all aligns correctly. Since the gene lists between 2 experiments may not be 100% identical, you will need to compare them side by side to make sure every gene row is identical. Remove any genes rows that are not shared among ALL samples until they all align.
Tip: Excel may try to convert the gene symbol column to dates automatically, which will cause issues. You can format the cells to be Text to prevent this, see here
3. Copy and paste the sample columns you want to keep in a new file. Alternatively, you could delete the sample columns you do not want to include.
Tip: Make sure to completely select and delete the entire columns or rows that you want to remove. If only the data is removed from the cells and not the entire row/column, it will cause issues when you try to export your file format from XLS to CSV and you may see blanks separated by commas in your new file
4. Export the file to a .CSV format and save it
Please reach out to Customer Support if you have any questions or concerns with this process and our team can help combine the data for you using Option 1 above.
Step 4: Create your new attribute file
Create a new combined attribute file for the new list of samples. You can copy and paste the relevant meta data from your existing Attribute files downloaded in Step 2, but make sure each sample has a value entered for every attribute.
We recommend that you add an extra attribute (such as 'Batch') to help track which experiment each sample is coming from, especially if the samples were run with varying sample prep kits. This will help you to better assess any batch like effects that may be present.
Step 5: Upload a new counts experiment
- Choose Design a New Experiment',
- Method= 'Sequence' and Experiment type= 'RNA-seq (Processed Counts)' or Small RNA-seq (Processed Counts)
- Processed Counts Type= Raw
Step 6: Share and unlock your analysis
- Wait for the experiment to finish processing to see the analysis unlock screen. Then share it with email@example.com (guide linked here) and let us know your Sample Reuse request is ready and we will send you the AU needed to unlock the new processed counts experiment.