The Cross Evaluation process is used to evaluate potential crosses using scoring code. Scoring code is a predictive model saved into a SAS or JSL file.
Cross Evaluation evaluates all crosses among rows of a data set containing biallelic numeric genotypes using scoring code for traits. It outputs a model-based mean, max, min, range, and standard deviation for each cross and trait. It optionally also simulates progeny from all crosses and produces the same summary statistics.
The Input Data Set must contain columns identifying the genotypes for each of the markers. Markers must be biallelic and defined using a numeric format. Genotypes must be coded as 0, 1, 2, or dot (.) (for missing value) before this data set can be input into this process.
The sampledata_numgeno.sas7bdat data set shown above is included in the Sample Data folder. The trait to be assessed and the biallelic markers (0 and 2 represent individuals homozygous for one of two possible alleles, 1represents the heterozygote) are indicated.
For detailed information about the files and data sets used or created by JMP Life Sciences software, see Files and Data Sets.
The second required file is one or more Scoring Code Files. These files are SAS program files that contain models that use measured attributes to either characterize or predict the value of an event. These models are developed on historical data where an event has been measured or inferred and are generated using one or more predictive modeling processes, such as Ridge Regression, for example. The models are then applied to new data for which the attributes are known, but the event has not yet occurred.
The sampledata_numgeno_rr_score.sas file is shown below:
This file consists of one line of SAS code (spread over multiple lines in the file) that can be incorporated into a SAS DATA step in which the generated coefficients are multiplied by the variables to produce predictors for each cross.
The output generated by this process is summarized in a Tabbed report. Refer to the Cross Evaluation output documentation for detailed descriptions and guides to interpreting your results.