hlaGenotypeEvaluator

• The software compares test HLA genotypes against reference HLA genotypes
• hlaGenotypeEvaluator assigned a score for each allele tested

Prerequisite:

• For Windows computers, we recommend installing the Git Bash terminal (https://git-scm.com/download/win). The Git Bash terminal allows HaplObserve to be run the same way as in Linux and Mac terminals

• Download and Install Java SE Development Kit (JDK - 1.8 or newer)

• Download the latest version of jar file from the release

• alternatively, download of the jar file can be made directly on a computer with this command:

• curl -LO "link to the released jar file"

Input files

1. Input files can be Histoimmunogenetics Markup Language (HML: https://bioinformatics.bethematchclinical.org/hla-resources/hml/) or csv file format.

• The file extension of HML can be either ".xml" or ".hml".
• The HML report can be generated using commercially available NGS HLA genotyping software, such as MiaFora (Immucor), TypeStream Visual (One Lambda) or HLA Twin (Omixon)
• The reference HML contains "reference" report, and test HML includes test report
• The software takes <baseDirectory> as an argument. The <baseDirectory> name can be anything, such as validation, but do not use space in <baseDirectory> name.
• HML files must be stored in "ref" and "test" directories. The software requires the following directory and file structures:
• “<baseDirectory>/ref/XXX.xml”, “<baseDirectory>/ref/XXX.hml” or “<baseDirectory>/ref/XXX.csv”
• Input reference file (XXX) should be stored in “ref” directory, and "ref" directory name must be used.
• “<baseDirectory>/test/YYY.xml”, “<baseDirectory>/test/YYY.hml” or “<baseDirectory>/test/YYY.csv”
• Input test file (YYY) should be stored in “test” directory, and "test" directory name must be used.
• Only one HML file should be stored in each "ref" and "test" directories.
• The software can be executed as follows:
• java -jar hlaGenotypeEvaluator-0.0.1-SNAPSHOT-jar-with-dependencies.jar <baseDirectory>
• The output file is generated in “/TestResult_today.csv". The file name will have today's date.

2. This software takes csv files as input files, and generates csv output file as follows:

• The first argument should be a file containing reference HLA genotypes (CSV file format)
• The second argument should be a file that contains test HLA genotypes to be evaluated (CSV file format)
• These files should be in the same format. The first column indicates "Sample ID", and the second column shows "GL String". These files should contain header in the first line
• The third argument should be provided to specify output file name. Date and ".csv" extension is added by the software
• The software can be executed as follows:
• java -jar hlaGenotypeEvaluator-0.0.1-SNAPSHOT-jar-with-dependencies.jar reference.csv test.csv testResult

Scoring description in Results file

• the software generates a result csv file
• Results consisted of a column for each allele and locus with 3 rows per sample.
• The first row represents the "Reference" HLA genotypes.
• The second row shows "Test" HLA genotypes.
• The third row contains "Score" for the comparison between the reference genotypes and the test genotypes for each locus
• The "Score" includes the following wording
• Identical: HLA genotypes are exactly identical
• Concordance: Both Test and Ref are concordant at least by two field assessment
• AmbResultConcordant: Ambiguity reported in the Test but not in the Reference
• AmbRefConcordant: Ambiguity reported in the Reference
• AmbRefAmbResultConcordant: Ambiguity reported in the Reference and Test
• UnresolvedNullAmbResultConcordant: Concordant, but Null allele was found in the Test ambiguity string
• Discordant: Indicates both alleles do not match at all

Reference

Quality control project of NGS HLA genotyping for the 17th International HLA and Immunogenetics Workshop [https://www.ncbi.nlm.nih.gov/pubmed/30738112]