Choose a linkage grouping method

Choose a method for assigning markers to linkage groups.

Grouping methods are described in the following table:

Method	Description
Interactive Hierarchical Clustering	This method launches the JMP hierarchical clustering platform and performs clustering on the recombination distance matrix . Use the Interactive Clustering Method drop-down menu to specify the clustering method that you want to use. You must also specify an Initial Number of Linkage Groups when using this method. Note : The default clustering method for hierarchical clustering, Average Linkage Clustering , is more robust for forming linkage groups comprised of markers that on average have low recombination instead of defining groups based on solely the minimum recombination between a pair of markers. An Initial Number of Linkage Groups must be specified, but this can be interactively changed to determine the number of linkage groups that you want to form based on the structure of the heat map of recombination fractions and segregation ratio distribution .
Automated Hierarchical Clustering	This method runs SAS PROC CLUSTER to cluster markers into linkage groups. Linkage group formation can be based on either a specified number of clusters, or based on a recombination fraction cutoff. Use the Automated Linkage Group Clustering Method drop-down menu to specify the clustering method that you want to use. The three available automated clustering routines define groups based on different assumptions of the behavior of recombination fractions. You must specify how linkage groups are to be defined when using this method. • The Average clustering method defines the distance between groups based on the average recombination rate between markers in different clusters. • Complete clustering defines groups based on the maximum recombination fraction between a marker in one cluster and a marker in another cluster. • Single linkage defines linkage groups based on the minimum recombination fraction between a marker in one cluster and a marker in another. Note : The default method for this process, Average , is more robust for forming linkage groups comprised of markers that on average have low recombination instead of defining groups based on solely the minimum recombination between a pair of markers.
Minimum Recombination Grouping	This method uses SAS PROC OPTMODEL to use a constrained minimum spanning tree algorithm on the pairwise recombination fractions for markers. You must specify a Grouping Recombination Fraction Threshold and a Minimum Number of Linkage Groups when using this method. This method is analogous to Single Linkage clustering where it connects markers with edges based on minimum pairwise recombination fractions. This method can be constrained to honor a skeleton framework of markers belonging to previously defined linkage groups and enables you to add new markers into linkage groups as determined by the Framework Linkage Group Variable (specified on the Annotation tab). Note : This method must be used for linkage grouping when a Framework Linkage Group Variable is specified.
Transitive Grouping 1	The transitive grouping method is implemented in the "groups" function from the OneMap R package and uses the property of genetic linkage, that is, if markers M1 and M2 are linked, and M2 is linked to marker M3 , then M1 is also linked to M3 . Therefore, markers M1 , M2 , and M3 are linked together in the same linkage group. Two parameters are used to define linkage between markers, the logarithm of the odds ratio ( LOD ) and the recombination fraction. If the LOD for linkage between two markers is larger than a predefined LOD threshold and these markers have a recombination fraction estimate smaller than a predefined recombination fraction threshold, then these markers are defined as linked. Otherwise, they are defined as unlinked. Note : You must specify a Grouping LOD Threshold and a Grouping Recombination Fraction Threshold when selecting this grouping method.

---

1

Note : This method is available in the Outcross Linkage Map Analysis JMP Add-In only. This Add-In is available for installation from C:\Program Files\SASHome\JMPGenomics\12\Genomics\OutCross directory or for download from the JMP File Exchange web page.

To Specify a Linkage Grouping Method:

Select the desired grouping method by clicking within the appropriate radio button.

For Additional Information

Refer to the SAS PROC CLUSTER documentation and the SAS PROC OPTMODEL documentation for more information.