Skip to content

Legacy Documentation - PHG Version 1

This section contains documentation for PHG version 1, which is no longer actively developed. It is preserved here for archival and historical reference only. If you are looking to use the Practical Haplotype Graph, please refer to the PHG v2 documentation, which reflects the current version of the software.

Create a kmer index

The KmerHashMapFromGraphPlugin is a tool that creates a kmer hash map with a specified value of k (in this case, k = 32). It can optionally save this map to a file. Let’s break down its functionality:

  1. Purpose

  2. The primary purpose of this plugin is to create a map that associates each kmer hash with the haplotypes in which it occurs.

  3. It helps discriminate between different haplotypes based on observed kmers.

  4. Steps

  5. The initial step involves finding a set of kmers that are observed in only one reference range.

  6. These kmers occur in at most (maxHaplotypes * number of haplotypes) haplotypes in that reference range.

  7. Once a diagnostic kmer set is identified, the plugin creates a map of kmer hash values to the list of haplotypes in which they occur.

  8. Options

  9. -maxHaplotypes: Specifies the maximum number of haplotypes as a proportion of total haplotypes for a kmer to be considered.

  10. -hashMask: Determines how the positions of a kmer hash are masked for pre-filtering hashes.
  11. -hashFilter: Defines which kmers are used based on their masked values (0 = A, 1 = C, 2 = G, 3 = T).
  12. -saveFile: Specifies the name and path of the file where the map will be saved.

  13. -configFile: Requires a config file containing database connection parameters.

  14. Example Command

./run_pipeline.pl -Xmx100G -debug -configParameters /myDir/myConfig.txt -HaplotypeGraphBuilderPlugin -methods myMethod -includeSequences true -endPlugin -KmerHashMapFromGraphPlugin -saveFile myFileName -endPlugin

Details

Remember to replace the placeholders (/myDir/myConfig.txt, myMethod, myFileName) with actual values relevant to your use case. Setting -Xmx100G should provide enough RAM for most databases. Expect this step to take a few hours. This tool is particularly useful for mapping reads during imputation. 😊 Typing the following command produces help information: ./run_pipeliine.pl -KmerHashMapFromGraphPlugin
This returns the following:

KmerHashMapFromGraphPlugin Description...
This plugin creates a kmer hash map with k = 32 and, optionally, saves it to a file. The kmer map is a map of kmer hash to hapid list that associates each kmer hash with haplotypes it occurs in. The first step in creating the map is to find a set of kmers that are observed in only one reference range and that occur in at most maxHaplotypes of them, so that the kmers can discriminate between haplotypes. Once a diagnostic kmer set is found, a map is created of kmer hash value to the list of haplotypes in which they occur. The map is then saved to a file for later use in mapping reads for imputation.

Usage: KmerHashMapFromGraphPlugin
-maxHaplotypes : If a kmer is in more than (maxHaplotypes * number of haplotypes) in a reference range it will not be used. (Default: 0.75)
-hashMask : The value used to mask the positions of a kmer hash used to pre-filter hashes. A value of 3 uses the last position of the kmer, and a value of 15 uses the last two positionns (Default: 3)
-hashFilter : A kmer is used when its masked value equals hashFilter, where 0 = A, 1 = C, 2 = G, and 3 = T. (Default: 1)
-saveFile : The name and path of the file to which the map will be saved.
-configFile : A config file containing database connection parameters. Either this value or a value for -configParameters is required.