DownloadLandData.pdf
From Array Suite Wiki
This page has been migrated, find the latest information here
Download Land Data For Samples, For Selected Genes (Local analyses)
The function allows users to query the Land database with a list of gene and download them as OmicData/Table to Local Analysis project.
Data Types
The available data types for download depend on the availability in the Land; descriptions for the data types can be found here.
Options
- Land: the land data to query input genes
- Project: the target ArrayStudio project to store downloaded Land data
- SampleSet: specify a collection of samples (sampleSet) or get all samples
- VariableSet: specify a collection of columns (variableSet) or get none/all columns from sample meta data. Selected metadata will be downloaded as a table in the project. By default, only PrimaryGrouping and SecondaryGrouping columns are attached as design in each generated OmicData.
- Choose one or more data type to download: choose the land data type to download, each one will be a OmicData/Table in ArrayStudio project
- Data format
- Original data: the original data values as stored in Land, such as RNA-Seq FRKM at transcript level and microarray data at probeset level.
- Gene level data: data values are summarized at gene level
- Matrix data: data values are organized as feature * Samples in "microarray" format
- Full meta data: the full sample meta data will be downloaded as a table in the project. By default, only PrimaryGrouping and SecondaryGrouping columns are attached as design in each generated OmicData.
- Input genes: a list of genes to query. Gene can also be loaded from GeneSet
- Output name: name prefix for download data objects
Output
Downloaded data will be listed in the targeted project. These omic data will be saved as in-file rather than in-memory. If majority of the data matrix are zeros, such as feature * sample matrix for fusion, mutation and junction, the Omic data will be saved as sparse matrix data. These improvements significantly reduce memory usage by sacrificing little performance (data loading speed).
Here is one screenshot:
