Researchers from the Children’s Cancer Institute in Australia have developed a powerful new tool called consHLA to improve how doctors analyze HLA genes.

HLA, or human leukocyte antigens, are proteins that help the immune system recognize what belongs in the body and what does not. Accurate HLA typing is essential in medicine, especially for organ transplantation, understanding autoimmune diseases, and guiding cancer immunotherapy.

Traditional HLA testing methods like serology and PCR are reliable, but they only give limited information. Modern sequencing technologies, such as whole genome sequencing and RNA sequencing, can provide much more detail. However, analyzing HLA genes with sequencing alone can sometimes be tricky.

To overcome this challenge, the team combined three types of sequencing data: germline whole genome sequencing, tumor whole genome sequencing, and tumor RNA sequencing. By integrating these datasets into the consHLA workflow, they were able to provide highly accurate and clinically useful HLA typing results.

consHLA workflow

A Alignment, conversion and HLA typing is performed for each independent NGS input. An HLA referenced index for the workflow was previously generated from IPD-IMGT/HLA reference sequences. WGS or RNA-seq paired-end FASTQ files are aligned using Bowtie2 against the inputted HLA reference index. Reads are converted to FASTQ files and inputted to HLA-HD for HLA typing. The resulting JSON file is then passed to the consensus parser. B The workflow takes as input three filtered paired-end FASTQ files from germline and tumour WGS and tumour RNA-seq. HLA calling performed using HLA-HD (see A) and the HLA outputs then parsed by an allele consensus algorithm; the independent platform HLA results are outputted in JSON format, with consensus HLA alleles for both all and clinically relevant HLA genes outputted in both JSON and text format. A consensus HLA report is then generated as a PDF, following formats used in standard clinical HLA typing reports

In testing with data from 86 high-risk pediatric cancer patients, consHLA matched gold-standard clinical results with nearly 98% accuracy. Importantly, the tool also detected HLA changes in tumors that could have clinical significance. To make it practical for healthcare, the researchers designed the workflow to automatically generate clear, clinician-friendly reports.

This new approach could improve precision medicine for cancer patients by providing doctors with faster, more accurate genetic information to guide treatment decisions.

Availability – Full code is available at https://github.com/CCICB/consHLA