‹‹ Back to SVS Home

Haplotype Tables

11.10 Haplotype Tables

When a haplotype block is currently selected, the Tables for Block # section will be enabled in the Haplotype Block Set item’s attributes pane.

Clicking this button will launch a the Haplotype Tables dialog which allows you to compute various haplotype tables from the set of markers that were defined as a block. The haplotype tables will provide estimates for the probabilities of each haplotype for every sample using a given haplotype estimation method.

At the top of the dialog is a summary of how many markers were selected in the current block. For the purpose of the resulting tables you must define the parameters of the estimation algorithms that are used by the selected tables. See How Haplotype Frequencies are Computed for more details on the haplotype estimation algorithms and their options.

If the current LD plot was created from a spreadsheet with a case/control variable set as the single dependent variable of the spreadsheet, you will have the option of selecting the subset of samples defined as cases or controls as the samples for the resulting tables.

The Per sample EM and Per sample CHM check-boxes produces tables of haplotype frequencies estimated using the EM or CHM algorithm respectively with samples along the rows and haplotypes along the column. Only haplotypes with overall frequencies greater than the frequency threshold will be shown.

A diplotype is a pair of haplotypes from a given patient – one haplotype comes from mom, and the other comes from dad. A patient has one and only one diplotype. However, as with haplotypes, the lack of knowledge of marker phase makes it necessary to estimate the probability a patient has a given diplotype for a given set of markers. The Per sample diplotype check-box creates a diplotype table which presents a list of all probable diplotypes for all patients, where probable is defined by the EM display threshold. If you wish to only display patients whose diplotype probabilities are over, say, 0.8, then set the haplotype frequency threshold to 0.8.

We see in the diplotype table view that often the diplotype can be assigned with very high probability. This table can be output to other data formats for further analysis, or potentially joined with an existing table of response variables, and then the haplotype or diplotype entries can be used as categorical variables to build models of the response in different haplotype or diplotype subpopulations.