HLAProfiler utilizes k-mer profiles to improve HLA calling accuracy for rare and common alleles in RNA-seq data

Buchkovich, Martin L.; Brown, Chad; Robasky, Kimberly; Chai, Shengjie; Westfall, Sharon; Vincent, Benjamin G.; Weimer, Eric T.; Powers, Jason

doi:10.1186/s13073-017-0473-6

Cited by 45 publications

(49 citation statements)

References 26 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…Although several HLA-typing tools for RNAseq data exist [23–25], they do not provide expression quantification with UMI counting. By using our custom pipeline we were able to determine HLA mRNA expression levels to the allele level.…”

Section: Discussionmentioning

confidence: 99%

“…Precise identification of HLA alleles from NGS data is challenging due to the high polymorphism and homologous nature of HLA genes leading often to ambiguous typing results. Several existing tools, such as seq2HLA[23], HLAforest[24], and HLAProfiler[25], have been developed to perform HLA typing from short RNA sequencing reads using the whole transcriptome data. Even though these tools enable accurate and comprehensive allele determination, they only accept data with a very low error rate and are designed merely for short-read Illumina data.…”

Section: Introductionmentioning

confidence: 99%

See 1 more Smart Citation

HLA RNAseq reveals high allele-specific variability in mRNA expression

Johansson

Koskela

et al. 2018

Preprint

View full text Add to dashboard Cite

18The HLA gene complex is the most important, single genetic factor in susceptibility to most 19 diseases with autoimmune or autoinflammatory origin and in transplantation matching. The majority of 20 the studies have focused on the huge allelic variation in these genes; only a few studies have explored 21 differences in expression levels of HLA alleles. To study the expression levels of HLA alleles more 22 systematically we utilised two different RNA sequencing methods. Illumina RNAseq has a high 23 sequencing accuracy and depth but is limited by the short read length, whereas Oxford Nanopore's 24 technology can sequence long templates, but has a poor accuracy. We studied allelic mRNA levels of 25 HLA class I and II alleles from peripheral blood samples of 50 healthy individuals. The results 26 demonstrate large differences in mRNA expression levels between HLA alleles. The method can be 27 applied to quantitate the expression differences of HLA alleles in various tissues and to evaluate the role 28 of this type of variation in transplantation matching and susceptibility to autoimmune diseases. 29 3 30 Author Summary 31 32Even though HLA is widely studied less is known of its allele-specific expression. Due to the pivotal role 33 of HLA in infection response, autoimmunity, and transplantation biology its expression surely must play 34 a part as well. In hematopoietic stem cell transplantation the challenge often is to find a suitable HLA-35 matched donor due to the high allelic variation. Classical HLA typing methods do not take into account 36 HLA allele-specific expression. However, differential allelic expression levels could be crucial in finding 37 permissive mismatches in order to save a patient's life. Additionally, differential HLA expression levels 38 can lead into beneficial impact in viral clearance but also undesirable effects in autoimmune diseases. To 39 study HLA expression we developed a novel RNAseq-based method to systematically characterize allele-40 specific expression levels of classical HLA genes. We tested our method in a set of 50 healthy individuals 41 and found differential expression levels between HLA alleles as well as interindividual variability at the 42 gene level. Since NGS is already well adopted in HLA research the next step could be to determine HLA 43 allele-specific expression in addition to HLA allelic variation and HLA-disease association studies in 44 various cells, tissues, and diseases. 45 54 reaction (PCR) and the mean fluorescence intensity (MFI).[3-10] The differential expression of HLA 55 alleles has been associated with immunologically mediated diseases, such as Crohn's disease [11] and 56 HIV [6,12], follicular lymphoma[7], and the outcome of HSCT through the risk of graft versus host 57 disease (GvHD)[8,9]. In fact, incompatibilities between the donor and the recipient in HSCT have made 58 the expression differences of HLA molecules an interesting target for finding permissive mismatches. 59 Although currently only the qualitative HLA typing is considered in dono...

show abstract

Section: Discussionmentioning

confidence: 99%

Section: Introductionmentioning

confidence: 99%

HLA RNAseq reveals high allele-specific variability in mRNA expression

Johansson

Koskela

et al. 2018

Preprint

View full text Add to dashboard Cite

show abstract

“…We ran arcasHLA on these samples IMGT/HLA v3.24.0, the version used by HLAProfiler [7]. This version was selected by Buchkovich, instead of the latest version at the time of HLAProfiler's development v3.26.0, to increase the number of partial alleles in the dataset to demonstrate the tool's ability to call partial alleles.…”

Section: Benchmark Dataset: 1000 Genomesmentioning

confidence: 99%

“…HLAProfiler [7] is the top competitor for arcasHLA, as it is able to genotype both class I and class II MHC with high accuracy. arcasHLA, however, effectively achieves an order of magnitude runtime improvement over HLAProfiler when mapped RNA-seq reads are readily available, as HLAProfiler does not provide support for and does not benefit from pre-aligned sample input.…”

Section: Runtime Analysesmentioning

confidence: 99%

“…Although a plethora of tools optimized for typing from WES and WGS fall into either category, most current tools for RNA sequencing, including seq2HLA [5], OptiType [33], PHLAT [3], are alignment-based. The latest RNA-dedicated HLA typer, HLAProfiler, takes a novel approach to graph-based alignment, breaking the HLA transcripts into kmers and constructing a taxonomic tree used to filter reads [7]. To find an individual's genotype, observed k-mers are compared to profiles built from simulated reads.…”

Section: Introductionmentioning

confidence: 99%

See 1 more Smart Citation

arcasHLA: high resolution HLA typing from RNA seq

Orenbuch

Filip

Comito

et al. 2018

Preprint

View full text Add to dashboard Cite

Motivation. Human leukocyte antigen (HLA) locus makes up the major compatibility complex (MHC) and plays a critical role in host response to disease, including cancers and autoimmune disorders. In the clinical setting, HLA typing is necessary for determining tissue compatibility. Recent improvements in the quality and accessibility of next-generation sequencing have made HLA typing from standard short-read data practical. However, this task remains challenging given the high level of polymorphism and homology between the HLA genes. HLA typing from RNA sequencing is further complicated by post-transcriptional splicing and bias due to amplification.Results. Here, we present arcasHLA: a fast and accurate in silico tool that infers HLA genotypes from RNA sequencing data. Our tool outperforms established tools on the gold-standard benchmark dataset for HLA typing in terms of both accuracy and speed, with an accuracy rate of 100% at two field precision for MHC class I genes, and over 99.7% for MHC class II. Importantly, arcasHLA takes as its input pre-aligned BAM files, and outputs three-field resolution for all HLA genes in less than 2 minutes. Finally, we discuss evaluate the performance of our tool on a new biological dataset of 447 single-end total RNA samples from nasopharyngeal swabs, and establish the applicability of arcasHLA in metatranscriptome studies.Availability. arcasHLA is available at https://github.com/RabadanLab/arcasHLA.

show abstract