How to Use Target Explorer

Weights: Target Explorer allows the user to define weights to compute a weighted sum of desired attributes and sort the genes by this score. Choosing to set the weights to the Hasan weights will automatically set all relevant column weights to the ones derived by Hasan et. al. (2006) for gene prioritization.

Score Computation: The score appears as the right-most column. (Only integer or float data is considered while computing score). Missing attribute values(when data is unavailable from experiments) are treated as zeros.

Sorting: Target Explorer also allows the user to sort based on each of the different attributes. The default sorting is based on the overall score for each gene (descending order). The user is allowed to choose a single primary and a single secondary key for sorting.

Normalization: Allows the user to normalize all the 'real' attributes in two different ways: standard normalize (data normalized to have a mean of 0 and a standard deviation of 1) and range normalize (data normalized to lie between 0 and 1)

Constraints: Allows user to select genes based on user defined constraints for each of the attributes. If multiple constraints are specified, they are implicitly treated as a conjunction (and). Examples of acceptable constraint expressions are as follows:

  • Threshold operations: >0, <0, =-1
  • Range definition: > 0 and < 1
  • Negation: not=0
  • String matching:
  • Missing data: empty