SessionExpression
There are two types of expression sessions:
- The base type is just called an Expression session, which is where you already have the required data files on your computer, and you want to upload them to the app.
- If you want to extract data from GEO, then use a GEO import session.
Expression sessions have a specified collection of registered files in order to constrain the data curation and loading process. Once the main files are loaded into the expression session, then you can pick from these main steps in the wizard:
The functions on this page allow you to change the expression matrix to prepare it for loading into the app. We can put the functions into groups:
- Backup, restore, rename expression matrix files
- Simple formatting or cleaning, such as converting from CSV to txt, removing header rows or descriptive columns. These steps are useful if you grab data from a GEO supplementary file and want to make some simple changes.
- Feature names. If you want to convert feature/gene identifiers to gene symbols, or select one row for each gene symbol.
- Sample IDs. A couple functions to deal with sample IDs
- Normalize or scale the values
Curating a gene expression dataset requires reading about the experiment and ensuring that the metadata is accurate, at the dataset level and at the sample level. This can be iterative in that new ways to clean or annotate the dataset or samples might need to be done after the initial loading of the data into the SQL part of the app. These are some common use cases:
- Edit series info in order to change the series name, default plot information (group, color, expression units, dimensions, orientation).
- Edit sample annotations to make the sample labels more concise or consistent, split annotations into different columns (or combine them).
Generally speaking, the calculations you perform fit into these two categories:
- Creating scores like differential expression metrics between two groups.
- All other calculations including: PCA, volcano plots, expression distribution boxplots, global correlation search, etc.
The original purpose of the expression session was to stage data so it could be loaded into the SQL part of the app. Many of the other things on this page were added later.