HLAQuant

Author: Austin Crinklaw

What is it?

HLAQuant is a pipeline that produces fast and accurate allele specific expression for HLA genes. This is done by quantifying on the peptide binding groove domain using Salmon with personalized sequences.

Requirements:

• Linux OS
• NCBI Blast+
• Salmon -- please ensure Salmon is on your PATH!
• Python 3+
  • Python packages: Pandas, BioPython

How to use:

Installation:

HLAQuant can be downloaded through PyPI using the following pip command.

pip install hlaquant

Input

HLAQuant takes two input files currently.

• File one (-hla) consists of a sample_id and then a list of alleles corresponding to that sample's typing (tab separated)
• File two (-fastq) consists of a sample_id and then a list of FASTQ files corresponding to that sample (paired-end or single-end) Examples of these inputs can be found under the 'test_data/' directory

Usage

• A list of parameters and their descriptions can be found with the -h flag

python -m HLAQuant -h

Output

The output will match that of Salmons. It consists of a tab separated file containing the transcript ID (in this case, a specific HLA allele), as well as the number of reads. TPMs can be ignored as they will be inaccurate since we are only quantifying over a few sequences.

How does it work?

• We first take the list of alleles and fetch the corresponding sequences from IMGT
• Next we extract the sequences corresponding to their groove domains from these sequences
• We build an index for quantification using these G-domain sequences
• We then perform quantification using this index

The paper outlining this method in detail can be found [....somewhere when it is published]

References:

This pipeline would be unable to work without Salmon

Patro, R., Duggal, G., Love, M. I., Irizarry, R. A., & Kingsford, C. (2017). Salmon provides fast and bias-aware quantification of transcript expression. Nature Methods.