Next-generation sequencing in pharmacogenomics - fit for clinical decision support?

INTRODUCTION

The technological advances of sequencing methods during the past 20 years have fuelled the generation of large amounts of sequencing data that comprise common variations, as well as millions of rare and personal variants that would not be identified by conventional genotyping. While comprehensive sequencing is technically feasible, its clinical utility for guiding personalized treatment decisions remains controversial.

AREAS COVERED

We discuss the opportunities and challenges of comprehensive sequencing compared to targeted genotyping for pharmacogenomic applications. Current pharmacogenomic sequencing panels are heterogeneous and clinical actionability of the included genes is not a major focus. We provide a current overview and critical discussion of how current studies utilize sequencing data either retrospectively from biobanks, databases or repurposed diagnostic sequencing, or prospectively using pharmacogenomic sequencing.

EXPERT OPINION

While sequencing-based pharmacogenomics has provided important insights into genetic variations underlying the safety and efficacy of a multitude pharmacological treatments, important hurdles for the clinical implementation of pharmacogenomic sequencing remain. We identify gaps in the interpretation of pharmacogenetic variants, technical challenges pertaining to complex loci and variant phasing, as well as unclear cost-effectiveness and incomplete reimbursement. It is critical to address these challenges in order to realize the promising prospects of pharmacogenomic sequencing.

ASCARIS: Positional feature annotation and protein structure-based representation of single amino acid variations

Background

Genomic variations may cause deleterious effects on protein functionality and perturb biological processes. Elucidating the effects of variations is critical for developing novel treatment strategies for diseases of genetic origin. Computational approaches have been aiding the work in this field by modeling and analyzing the mutational landscape. However, new approaches are required, especially for accurate representation and data-centric analysis of sequence variations.

Method

In this study, we propose ASCARIS (Annotation and StruCture-bAsed RepresentatIon of Single amino acid variations), a method for the featurization (i.e., quantitative representation) of single amino acid variations (SAVs), which could be used for a variety of purposes, such as predicting their functional effects or building multi-omics-based integrative models. ASCARIS utilizes the direct and spatial correspondence between the location of the SAV on the sequence/structure and 30 different types of positional feature annotations (e.g., active/lipidation/glycosylation sites; calcium/metal/DNA binding, inter/transmembrane regions, etc.), along with structural features and physicochemical properties. The main novelty of this method lies in constructing reusable numerical representations of SAVs via functional annotations.

Results

We statistically analyzed the relationship between these features and the consequences of variations and found that each carries information in this regard. To investigate potential applications of ASCARIS, we trained variant effect prediction models that utilize our SAV representations as input. We carried out an ablation study and a comparison against the state-of-the-art methods and observed that ASCARIS has a competing and complementary performance against widely-used predictors. ASCARIS can be used alone or in combination with other approaches to represent SAVs from a functional perspective. ASCARIS is available as a programmatic tool at https://github.com/HUBioDataLab/ASCARIS and as a web-service at https://huggingface.co/spaces/HUBioDataLab/ASCARIS.

Structure-Informed Protein Language Models are Robust Predictors for Variant Effects

Predicting protein variant effects through machine learning is often challenged by the scarcity of experimentally measured effect labels. Recently, protein language models (pLMs) emerge as zero-shot predictors without the need of effect labels, by modeling the evolutionary distribution of functional protein sequences. However, biological contexts important to variant effects are implicitly modeled and effectively marginalized. By assessing the sequence awareness and the structure awareness of pLMs, we find that their improvements often correlate with better variant effect prediction but their tradeoff can present a barrier as observed in over-finetuning to specific family sequences. We introduce a framework of structure-informed pLMs (SI-pLMs) to inject protein structural contexts purposely and controllably, by extending masked sequence denoising in conventional pLMs to cross-modality denoising. Our SI-pLMs are applicable to revising any sequence-only pLMs through model architecture and training objectives. They do not require structure data as model inputs for variant effect prediction and only use structures as context provider and model regularizer during training. Numerical results over deep mutagenesis scanning benchmarks show that our SI-pLMs, despite relatively compact sizes, are robustly top performers against competing methods including other pLMs, regardless of the target protein family's evolutionary information content or the tendency to overfitting / over-finetuning. Learned distributions in structural contexts could enhance sequence distributions in predicting variant effects. Ablation studies reveal major contributing factors and analyses of sequence embeddings provide further insights. The data and scripts are available at https://github.com/Stephen2526/Structure-informed_PLM.git.

Prevalence of DDC genotypes in patients with aromatic L-amino acid decarboxylase (AADC) deficiency and in silico prediction of structural protein changes

Aromatic L-amino acid decarboxylase (AADC) deficiency is a rare autosomal recessive genetic disorder affecting the biosynthesis of dopamine, a precursor of both norepinephrine and epinephrine, and serotonin. Diagnosis is based on the analysis of CSF or plasma metabolites, AADC activity in plasma and genetic testing for variants in the DDC gene. The exact prevalence of AADC deficiency, the number of patients, and the variant and genotype prevalence are not known. Here, we present the DDC variant (n = 143) and genotype (n = 151) prevalence of 348 patients with AADC deficiency, 121 of whom were previously not reported. In addition, we report 26 new DDC variants, classify them according to the ACMG/AMP/ACGS recommendations for pathogenicity and score them based on the predicted structural effect. The splice variant c.714+4A>T, with a founder effect in Taiwan and China, was the most common variant (allele frequency = 32.4%), and c.[714+4A>T];[714+4A>T] was the most common genotype (genotype frequency = 21.3%). Approximately 90% of genotypes had variants classified as pathogenic or likely pathogenic, while 7% had one VUS allele and 3% had two VUS alleles. Only one benign variant was reported. Homozygous and compound heterozygous genotypes were interpreted in terms of AADC protein and categorized as: i) devoid of full-length AADC, ii) bearing one type of AADC homodimeric variant or iii) producing an AADC protein population composed of two homodimeric and one heterodimeric variant. Based on structural features, a score was attributed for all homodimers, and a tentative prediction was advanced for the heterodimer. Almost all AADC protein variants were pathogenic or likely pathogenic.

SUsPECT: a pipeline for variant effect prediction based on custom long-read transcriptomes for improved clinical variant annotation

Our incomplete knowledge of the human transcriptome impairs the detection of disease-causing variants, in particular if they affect transcripts only expressed under certain conditions. These transcripts are often lacking from reference transcript sets, such as Ensembl/GENCODE and RefSeq, and could be relevant for establishing genetic diagnoses. We present SUsPECT (Solving Unsolved Patient Exomes/gEnomes using Custom Transcriptomes), a pipeline based on the Ensembl Variant Effect Predictor (VEP) to predict variant impact on custom transcript sets, such as those generated by long-read RNA-sequencing, for downstream prioritization. Our pipeline predicts the functional consequence and likely deleteriousness scores for missense variants in the context of novel open reading frames predicted from any transcriptome. We demonstrate the utility of SUsPECT by uncovering potential mutational mechanisms of pathogenic variants in ClinVar that are not predicted to be pathogenic using the reference transcript annotation. In further support of SUsPECT's utility, we identified an enrichment of immune-related variants predicted to have a more severe molecular consequence when annotating with a newly generated transcriptome from stimulated immune cells instead of the reference transcriptome. Our pipeline outputs crucial information for further prioritization of potentially disease-causing variants for any disease and will become increasingly useful as more long-read RNA sequencing datasets become available.

High-throughput deep learning variant effect prediction with Sequence UNET

Understanding coding mutations is important for many applications in biology and medicine but the vast mutation space makes comprehensive experimental characterisation impossible. Current predictors are often computationally intensive and difficult to scale, including recent deep learning models. We introduce Sequence UNET, a highly scalable deep learning architecture that classifies and predicts variant frequency from sequence alone using multi-scale representations from a fully convolutional compression/expansion architecture. It achieves comparable pathogenicity prediction to recent methods. We demonstrate scalability by analysing 8.3B variants in 904,134 proteins detected through large-scale proteomics. Sequence UNET runs on modest hardware with a simple Python package.

aRgus: Multilevel visualization of non-synonymous single nucleotide variants & advanced pathogenicity score modeling for genetic vulnerability assessment

The widespread use of high-throughput sequencing techniques is leading to a rapidly increasing number of disease-associated variants of unknown significance and candidate genes. Integration of knowledge concerning their genetic, protein as well as functional and conservational aspects is necessary for an exhaustive assessment of their relevance and for prioritization of further clinical and functional studies investigating their role in human disease. To collect the necessary information, a multitude of different databases has to be accessed and data extraction from the original sources commonly is not user-friendly and requires advanced bioinformatics skills. This leads to a decreased data accessibility for a relevant number of potential users such as clinicians, geneticist, and clinical researchers. Here, we present aRgus (https://argus.urz.uni-heidelberg.de/), a standalone webtool for simple extraction and intuitive visualization of multi-layered gene, protein, variant, and variant effect prediction data. aRgus provides interactive exploitation of these data within seconds for any known gene of the human genome. In contrast to existing online platforms for compilation of variant data, aRgus complements visualization of chromosomal exon-intron structure and protein domain annotation with ClinVar and gnomAD variant distributions as well as position-specific variant effect prediction score modeling. aRgus thereby enables timely assessment of protein regions vulnerable to variation with single amino acid resolution and provides numerous applications in variant and protein domain interpretation as well as in the design of in vitro experiments.

DiMSum: an error model and pipeline for analyzing deep mutational scanning data and diagnosing common experimental pathologies

Deep mutational scanning (DMS) enables multiplexed measurement of the effects of thousands of variants of proteins, RNAs, and regulatory elements. Here, we present a customizable pipeline, DiMSum, that represents an end-to-end solution for obtaining variant fitness and error estimates from raw sequencing data. A key innovation of DiMSum is the use of an interpretable error model that captures the main sources of variability arising in DMS workflows, outperforming previous methods. DiMSum is available as an R/Bioconda package and provides summary reports to help researchers diagnose common DMS pathologies and take remedial steps in their analyses.

Variant effect predictions capture some aspects of deep mutational scanning experiments

BACKGROUND

Deep mutational scanning (DMS) studies exploit the mutational landscape of sequence variation by systematically and comprehensively assaying the effect of single amino acid variants (SAVs; also referred to as missense mutations, or non-synonymous Single Nucleotide Variants - missense SNVs or nsSNVs) for particular proteins. We assembled SAV annotations from 22 different DMS experiments and normalized the effect scores to evaluate variant effect prediction methods. Three trained on traditional variant effect data (PolyPhen-2, SIFT, SNAP2), a regression method optimized on DMS data (Envision), and a naïve prediction using conservation information from homologs.

RESULTS

On a set of 32,981 SAVs, all methods captured some aspects of the experimental effect scores, albeit not the same. Traditional methods such as SNAP2 correlated slightly more with measurements and better classified binary states (effect or neutral). Envision appeared to better estimate the precise degree of effect. Most surprising was that the simple naïve conservation approach using PSI-BLAST in many cases outperformed other methods. All methods captured beneficial effects (gain-of-function) significantly worse than deleterious (loss-of-function). For the few proteins with multiple independent experimental measurements, experiments differed substantially, but agreed more with each other than with predictions.

CONCLUSIONS

DMS provides a new powerful experimental means of understanding the dynamics of the protein sequence space. As always, promising new beginnings have to overcome challenges. While our results demonstrated that DMS will be crucial to improve variant effect prediction methods, data diversity hindered simplification and generalization.

Variant effect prediction tools assessed using independent, functional assay-based datasets: implications for discovery and diagnostics

Background

Genetic variant effect prediction algorithms are used extensively in clinical genomics and research to determine the likely consequences of amino acid substitutions on protein function. It is vital that we better understand their accuracies and limitations because published performance metrics are confounded by serious problems of circularity and error propagation. Here, we derive three independent, functionally determined human mutation datasets, UniFun, BRCA1-DMS and TP53-TA, and employ them, alongside previously described datasets, to assess the pre-eminent variant effect prediction tools.

Results

Apparent accuracies of variant effect prediction tools were influenced significantly by the benchmarking dataset. Benchmarking with the assay-determined datasets UniFun and BRCA1-DMS yielded areas under the receiver operating characteristic curves in the modest ranges of 0.52 to 0.63 and 0.54 to 0.75, respectively, considerably lower than observed for other, potentially more conflicted datasets.

Conclusions

These results raise concerns about how such algorithms should be employed, particularly in a clinical setting. Contemporary variant effect prediction tools are unlikely to be as accurate at the general prediction of functional impacts on proteins as reported prior. Use of functional assay-based datasets that avoid prior dependencies promises to be valuable for the ongoing development and accurate benchmarking of such tools.

Electronic supplementary material

The online version of this article (doi:10.1186/s40246-017-0104-8) contains supplementary material, which is available to authorized users.

GTB - an online genome tolerance browser

Background

Accurate methods capable of predicting the impact of single nucleotide variants (SNVs) are assuming ever increasing importance. There exists a plethora of in silico algorithms designed to help identify and prioritize SNVs across the human genome for further investigation. However, no tool exists to visualize the predicted tolerance of the genome to mutation, or the similarities between these methods.

Results

We present the Genome Tolerance Browser (GTB, http://gtb.biocompute.org.uk): an online genome browser for visualizing the predicted tolerance of the genome to mutation. The server summarizes several in silico prediction algorithms and conservation scores: including 13 genome-wide prediction algorithms and conservation scores, 12 non-synonymous prediction algorithms and four cancer-specific algorithms.

Conclusion

The GTB enables users to visualize the similarities and differences between several prediction algorithms and to upload their own data as additional tracks; thereby facilitating the rapid identification of potential regions of interest.

Electronic supplementary material

The online version of this article (doi:10.1186/s12859-016-1436-4) contains supplementary material, which is available to authorized users.