Skip to content

Latest commit

 

History

History
13 lines (12 loc) · 1.5 KB

hd-params.md

File metadata and controls

13 lines (12 loc) · 1.5 KB
Raw

High dimensional data parameters

  • DATA_FILE (Mandatory) Points to the HD data file. Preferred over DATA_FILE_PREFIX.
  • DATA_FILE_PREFIX (Deprecated)
  • DATA_TYPE (Mandatory) The current state of your data values. Can be R (raw values) or L (log-transformed values).
  • LOG_BASE Default: 2. In case you provide raw values, this is the base used for doing the log transformation. Currently only supports 2.
  • SRC_LOG_BASE (Mandatory if DATA_TYPE=L) The log base that was used for transforming the data to their current values. Has to be specified only with DATA_TYPE=L. This will be used to reconstruct the raw intensity values.
  • MAP_FILENAME (Mandatory) Name of the subject-sample mapping file.
  • ALLOW_MISSING_ANNOTATIONS Default: N. Y for yes, N for no. Whether the job should be allowed to continue when the data set doesn't provide data for all the annotations.
  • SKIP_UNMAPPED_DATA Default: N. If Y, it ignores data points of samples that are not present in the subject-sample mapping file. Otherwise (N) gives an error for such data points.
  • ZERO_MEANS_NO_INFO Default: N. If Y, raw values that equal 0 will be filtered out. Otherwise (N), they will be inserted into the database. The flag applies to most of HD data types. It does not affect CNV (aCGH) data. For RNASeq read count data the check on zeros happens based on normalized read count.