Each entry in this table contains information about the statistic used and the result of a particular integrity statistic run involving a trusted set of values in RAD and a newly added set of corresponding values, to catch possible inconsistencies with the loading of the latter. An example is to compute the correlation between raw and normalized log2 ratios in a given 2-channel array experiment to check that the normalized values were properly loaded.
|INTEGRITY_STATISTIC_ID||no||NUMBER(8,0)||primary key Edit|
|STATISTIC_METHOD||no||STRING(200)||This describes which statistic was used, e.g. correlation. Edit|
|TRUSTED_INPUT_FORMULA||STRING(200)||This explains how the trusted inputs were combined prior to be passed to the statistical computation, in terms of their IntegrityStatInput.row_designations. For example, if the integrity statistic computes the correlation between raw and normalized log2 ratios in a given 2-channel array experiment, for each element on the array there are 3 inputs: the raw value of the numerator (trusted), say with row_designation n, the raw value of the denominator (trusted), say with raw_designation d, and the value of the normalized log2 ratio. The trusted_input_formula in this case is log2(n/d). Edit|
|VALUE||no||STRING(100)||The value of the resulting statistic. Edit|