Skip to content

Latest commit

 

History

History
77 lines (46 loc) · 3.54 KB

HLA-ASM.md

File metadata and controls

77 lines (46 loc) · 3.54 KB
Raw

HLA*ASM

Basics

HLA*ASM can find HLA gene exon coordinates in long read-based assemblies and carry out HLA typing at G group resolution.

It can also carry out a comparison to known sample HLA types, which can be a useful assembly quality criterion.

Installing HLA*ASM

HLA*ASM is part of HLA*LA.

Prerequisites

bwa >= 0.7.12 nucmer (part of MUMmer3.23) List::MoreUtils (Perl module) Text::LevenshteinX (Perl module) Bio::DB::HTS (Perl module)

Installing the Perl modules

Installing Text::LevenshteinXS and List::MoreUtils should be straighforward and be as simple as perl -MCPAN -e shell, followed by install Text::LevenshteinXS and install List::MoreUtils.

Installing Bio::DB::HTS can be a bit more tricky. You can try the CPAN approach, but might want to refer to the GitHub repository for more information.

Download the data package

Download the data package (https://www.dropbox.com/s/mnkig0fhaym43m0/reference_HLA_ASM.tar.gz, md5sum 4e85ddb1ca1130710815f9c0e5e52e22) and install it into the main directory of HLA*LA:

cd HLA-LA/src
wget https://www.dropbox.com/s/mnkig0fhaym43m0/reference_HLA_ASM.tar.gz
tar -xvzf reference_HLA_ASM.tar.gz

Modifying paths.ini

Like HLALA, HLAASM uses the path.ini mechanism to find its dependencies. Refer to README.md for details. HLA*ASM has a separate working directory.

Running HLA*ASM

HLA-ASM.pl --assembly_fasta /path/to/assembly.fasta --sampleID MySample --truth /path/to/HLAreference.txt

Parameters

  • --assembly_fasta: Path to a single combined FASTA file of the assembly.
  • --sampleID: A unique sample ID. Calls with different sample IDs can be run in parallel.
  • --truth: An HLA truth file (optional). For format see the examples in the reference_HLA_ASM directory.

All output goes by default into output_HLA_ASM/$mySampleID (where $mySampleID is the value if --sampleID). Use a unique sample ID for each sample.

Interpreting the output from HLA*LA

The main output file is output_HLA_ASM/$mySampleID/summary.txt.

Columns:

  • contigID: contig ID.
  • locus: HLA locus.
  • calledGenotypes: called HLA genotypes, G group resolution.
  • components: exons used for determination of the HLA genotype.
  • editDistance_calledGenotypes_assembly: edit distance between the called HLA genotypes and the assembly.
  • minEditDistance_assembly_truth: minimum edit distance between the assembly and any of the provided true HLA types for the locus.
  • minEditDistance_calledGenotype_truth: minimum edit distance between the called HLA genotypes and any of the provided true HLA types for the locus.
  • minEditDistance_assembly_truth_whichAlleles: alleles (from the truth set) underlying the field minEditDistance_assembly_truth.
  • minEditDistance_calledGenotype_truth_whichAlleles: allele pairs (called HLA genotypes / truth type) underlying the field minEditDistance_assembly_truth.

We also generate additional *.tab files in output_HLA_ASM/$mySampleID/temp_alignments_separate, the most interesting of which is genePositions.tab. It contains the positions of found genes and exons. These can be used to determine gene presence/absence from the assembly or can serve as the basis for higher-resolution typing.

Citing HLA*ASM

For the time being, please cite the original HLA*PRG paper.