Variant Effect Predictor Research Articles

Spectrum of DDC variants causing aromatic l-amino acid decarboxylase (AADC) deficiency and pathogenicity interpretation using ACMG-AMP/ACGS recommendations.

Pathogenic variants in dopa decarboxylase (DDC), the gene encoding the aromatic l-amino acid decarboxylase (AADC) enzyme, lead to a severe deficiency of neurotransmitters, resulting in neurological, neuromuscular, and behavioral manifestations clinically characterized by developmental delays, oculogyric crises, dystonia, and severe neurologic dysfunction in infancy. Historically, therapy has been aimed at compensating for neurotransmitter abnormalities, but response to pharmacologic therapy varies, and in most cases, the therapy shows little or no benefit. A novel human DDC gene therapy was recently approved in the European Union that targets the underlying genetic cause of the disorder, providing a new treatment option for patients with AADC deficiency. However, the applicability of human DDC gene therapy depends on the ability of laboratories and clinicians to interpret the results of genetic testing accurately enough to diagnose the patient. An accurate interpretation of genetic variants depends in turn on expert-guided curation of locus-specific databases. The purpose of this research was to identify previously uncharacterized DDC variants that are of pathologic significance in AADC deficiency as well as characterize and curate variants of unknown significance (VUSs) to further advance the diagnostic accuracy of genetic testing for this condition. DDC variants were identified using existing databases and the literature. The pathogenicity of the variants was classified using modified American College of Medical Genetics and Genomics/Association for Molecular Pathology/Association for Clinical Genomic Science (ACMG-AMP/ACGS) criteria. To improve the current variant interpretation recommendations, in silico variant interpretation tools were combined with structural 3D modeling of protein variants and applied comparative analysis to predict the impact of the variant on protein function. A total of 422 variants were identified (http://biopku.org/home/pnddb.asp). Variants were identified on nearly all introns and exons of the DDC gene, as well as the 3' and 5' untranslated regions. The largest percentage of the identified variants (48%) were classified as missense variants. The molecular effects of these missense variants were then predicted, and the pathogenicity of each was classified using a number of variant effect predictors. Using ACMG-AMP/ACGS criteria, 7% of variants were classified as pathogenic, 32% as likely pathogenic, 58% as VUSs of varying subclassifications, 1% as likely benign, and 1% as benign. For 101 out of 108 reported genotypes, at least one allele was classified as pathogenic or likely pathogenic. In silico variant pathogenicity interpretation tools, combined with structural 3D modeling of variant proteins and applied comparative analysis, have improved the current DDC variant interpretation recommendations, particularly of VUSs.

In silico interpretation of variants identifies novel causes of inherited eye disease in the 100 000 genomes project

AbstractInherited retinal dystrophy (IRD) describes a widely heterogeneous spectrum of disorders with over 270 genes implicated to date across all inheritance patterns and an estimated prevalence of around 1 in 2000 people. Genetic testing for IRD had evolved dramatically since the discovery of the first pathogenic variant in 1990: the advent of massively parallel sequencing means that the more than half of patients are now able to receive a molecular diagnosis although there is still significant missing heritability (up to ~40% of patients). Novel genes, non‐coding variants, structural variants/copy number variants, novel gene/disease associations, non‐Mendelian inheritance and non‐genetic disease are some of the mechanisms thought to account for the missing heritability of rare disease. This presentation will focus on non‐coding variant analysis.Whole genome sequencing (WGS) enables the capture and interrogation of all variants covering up to approximately 95% of the genome, and advances in genomic analysis technology, genome and variant databases, bioinformatics and in silico tools mean that nowadays variant prioritization pipelines are able to filter the huge variant numbers leaving only a handful of candidate pathogenic coding variants per patient.For the missing heritability in patients whose phenotype cannot be fully explained by this approach, however, non‐coding variant analysis is more challenging. Many more variants exist outside of coding exons and our understanding of their effect on transcripts and proteins is limited. Functional assays, especially for retina specific genes may be complex and difficult to establish, so accurate variant effect predictions are critical to enable molecular geneticists and clinical laboratories to maximize their ability to provide a definitive molecular diagnosis.This presentation will focus on the analysis of non‐coding variants in IRD genes in a cohort of patients who underwent WGS in the UK's 100 000 genomes project. Key to this study, patients were identified for study following multidisciplinary team discussion to highlight the most likely gene to be implicated in their disease. Identification of candidate non‐coding variants, filtering and prioritization were performed to identify the most compelling variant, followed by molecular investigation and confirmation of the predicted effect. Exemplar case studies will demonstrate the power of this analysis pipeline and accuracy of in silico predictions.

Functional genomics of OCTN2 variants informs protein-specific variant effect predictor for Carnitine Transporter Deficiency

Genetic variants in SLC22A5, encoding the membrane carnitine transporter OCTN2, cause the rare metabolic disorder Carnitine Transporter Deficiency (CTD). CTD is potentially lethal but actionable if detected early, with confirmatory diagnosis involving sequencing of SLC22A5. Interpretation of missense variants of uncertain significance (VUSs) is a major challenge. In this study, we sought to characterize the largest set to date (n = 150) of OCTN2 variants identified in diverse ancestral populations, with the goals of furthering our understanding of the mechanisms leading to OCTN2 loss-of-function (LOF) and creating a protein-specific variant effect prediction model for OCTN2 function. Uptake assays with 14C-carnitine revealed that 105 variants (70%) significantly reduced transport of carnitine compared to wild-type OCTN2, and 37 variants (25%) severely reduced function to less than 20%. All ancestral populations harbored LOF variants; 62% of green fluorescent protein (GFP)-tagged variants impaired OCTN2 localization to the plasma membrane of human embryonic kidney (HEK293T) cells, and subcellular localization significantly associated with function, revealing a major LOF mechanism of interest for CTD. With these data, we trained a model to classify variants as functional (>20% function) or LOF (<20% function). Our model outperformed existing state-of-the-art methods as evaluated by multiple performance metrics, with mean area under the receiver operating characteristic curve (AUROC) of 0.895 ± 0.025. In summary, in this study we generated a rich dataset of OCTN2 variant function and localization, revealed important disease-causing mechanisms, and improved upon machine learning-based prediction of OCTN2 variant function to aid in variant interpretation in the diagnosis and treatment of CTD.

Ensembl 2023.

Ensembl (https://www.ensembl.org) has produced high-quality genomic resources for vertebrates and model organisms for more than twenty years. During that time, our resources, services and tools have continually evolved in line with both the publicly available genome data and the downstream research and applications that utilise the Ensembl platform. In recent years we have witnessed a dramatic shift in the genomic landscape. There has been a large increase in the number of high-quality reference genomes through global biodiversity initiatives. In parallel, there have been major advances towards pangenome representations of higher species, where many alternative genome assemblies representing different breeds, cultivars, strains and haplotypes are now available. In order to support these efforts and accelerate downstream research, it is our goal at Ensembl to create high-quality annotations, tools and services for species across the tree of life. Here, we report our resources for popular reference genomes, the dramatic growth of our annotations (including haplotypes from the first human pangenome graphs), updates to the Ensembl Variant Effect Predictor (VEP), interactive protein structure predictions from AlphaFold DB, and the beta release of our new website.

Activating IGF1R hotspot non-frameshift insertions define a novel, potentially targetable molecular subtype of adenoid cystic carcinoma

AbstractActivation of the tyrosine kinase receptor IGF1R is targetable with existing tyrosine kinase inhibitors (TKIs) and monoclonal antibodies, but mutations in IGF1R have not been systematically characterized. Pan-cancer analysis of 326,911 tumors identified two distinct, activating non-frameshift insertion hotspots in IGF1R, which were significantly enriched in adenoid cystic carcinomas (ACCs). IGF1R alterations from 326,911 subjects were analyzed by variant effect prediction class, position within the gene, and cancer type. 6502 (2.0%) samples harbored one or more alterations in IGF1R. Two regions were enriched for non-frameshift insertions: codons 663–666 at the hinge region of the fibronectin type 3 domain and codons 1034–1049 in the tyrosine kinase domain. Hotspot insertions were highly enriched in ACCs (27.3-fold higher than in the remainder of the pan-cancer dataset; P = 2.3 × 10−17). Among salivary gland tumors, IGF1R hotspot insertions were entirely specific to ACCs. IGF1R alterations were most often mutually exclusive with other ACC drivers (9/15, 60%). Tumors with non-frameshift hotspot IGF1R insertions represent a novel, potentially targetable subtype of ACC. Additional studies are needed to determine whether these patients respond to existing IGF1R inhibitors.

Understanding molecular mechanisms and predicting phenotypic effects of pathogenic tubulin mutations.

Cells rely heavily on microtubules for several processes, including cell division and molecular trafficking. Mutations in the different tubulin-α and -β proteins that comprise microtubules have been associated with various diseases and are often dominant, sporadic and congenital. While the earliest reported tubulin mutations affect neurodevelopment, mutations are also associated with other disorders such as bleeding disorders and infertility. We performed a systematic survey of tubulin mutations across all isotypes in order to improve our understanding of how they cause disease, and increase our ability to predict their phenotypic effects. Both protein structural analyses and computational variant effect predictors were very limited in their utility for differentiating between pathogenic and benign mutations. This was even worse for those genes associated with non-neurodevelopmental disorders. We selected tubulin-α and -β disease mutations that were most poorly predicted for experimental characterisation. These mutants co-localise to the mitotic spindle in HeLa cells, suggesting they may exert dominant-negative effects by altering microtubule properties. Our results show that tubulin mutations represent a blind spot for current computational approaches, being much more poorly predicted than mutations in most human disease genes. We suggest that this is likely due to their strong association with dominant-negative and gain-of-function mechanisms.

Estimating diagnostic noise in panel-based genomic analysis.

Gene panels with a series of strict variant filtering rules are often used for clinical analysis of exomes and genomes. Panel sizes vary, affecting the test's sensitivity and specificity. We investigated the background rate of candidate variants in a population setting using gene panels developed to diagnose a range of heterogeneous monogenic diseases. We used the Gene2Phenotype database with the Variant Effect Predictor plugin to identify rare nonsynonymous variants in exome sequence data from 200,643 individuals in UK Biobank. We evaluated 5 clinically curated gene panels of varying sizes (50-1700 genes). Bigger gene panels resulted in more prioritized variants, varying from an average of approximately 0.3 to 3.5 variants per person. The number of individuals with prioritized variants varied linearly with coding sequence length for monoallelic genes (∼300 individuals per 1000 base pairs) and quadratically for biallelic genes, with notable outliers. Although large gene panels may be the best strategy to maximize diagnostic yield in genetically heterogeneous diseases, they frequently prioritize likely benign variants requiring follow up. Most individuals have ≥1 rare nonsynonymous variant in panels containing >500 disease genes. Extreme caution should be applied when interpreting candidate variants, particularly in the absence of relevant phenotypes.

Gene-specific machine learning model to predict the pathogenicity of BRCA2 variants.

Background: Existing BRCA2-specific variant pathogenicity prediction algorithms focus on the prediction of the functional impact of a subtype of variants alone. General variant effect predictors are applicable to all subtypes, but are trained on putative benign and pathogenic variants and do not account for gene-specific information, such as hotspots of pathogenic variants. Local, gene-specific information have been shown to aid variant pathogenicity prediction; therefore, our aim was to develop a BRCA2-specific machine learning model to predict pathogenicity of all types of BRCA2 variants. Methods: We developed an XGBoost-based machine learning model to predict pathogenicity of BRCA2 variants. The model utilizes general variant information such as position, frequency, and consequence for the canonical BRCA2 transcript, as well as deleteriousness prediction scores from several tools. We trained the model on 80% of the expert reviewed variants by the Evidence-Based Network for the Interpretation of Germline Mutant Alleles (ENIGMA) consortium and tested its performance on the remaining 20%, as well as on an independent set of variants of uncertain significance with experimentally determined functional scores. Results: The novel gene-specific model predicted the pathogenicity of ENIGMA BRCA2 variants with an accuracy of 99.9%. The model also performed excellently on predicting the functional consequence of the independent set of variants (accuracy was up to 91.3%). Conclusion: This new, gene-specific model is an accurate method for interpreting the pathogenicity of variants in the BRCA2 gene. It is a valuable addition for variant classification and can prioritize unreviewed variants for functional analysis or expert review.

Association analysis of risk genes identified by SCHEMA with schizophrenia in the Chinese Han population.

Schizophrenia is a chronic brain disorder. Previously, the Schizophrenia Exome Sequencing Meta-analysis consortium identified 10 highest risk genes related to schizophrenia. This study aimed to analyze the relationship between the 10 highest risk genes identified by the SCHEMA and schizophrenia in a Chinese population. A total of 225 variants in 10 genes were screened in a Chinese population of 6836 using a customized array. All variants were annotated through the Variant Effect Predictor tool, and the functional impacts of missense variants were assessed based on sorting intolerant from tolerant and PolyPhen-2 scores. The SHEsisPlus tool was used to analyze the association between risk genes and schizophrenia at the locus and gene levels. At the locus level, no missense variants significantly related to schizophrenia were found, but we detected three missense variants that appeared only in cases, including TRIO p. Arg1185Gln, RB1CC1 p. Arg1514Cys, and HERC1 p. Val4517Leu. At the gene level, five genes (TRIO, RB1CC1, HERC1, GRIN2A, and CACAN1G) with more than one variant analyzed were kept for the gene-level association analysis. Only the association between RB1CC1 and schizophrenia reached a significant level (OR = 1.634; 95% CI, 1.062-2.516; P = 0.025). In this study, we determined that RB1CC1 might be a risk gene for schizophrenia in the Chinese population. Our results provide new evidence for recognizing the correlation of these risk genes with the Chinese schizophrenia population.

Identifying interpretable gene-biomarker associations with functionally informed kernel-based tests in 190,000 exomes

Here we present an exome-wide rare genetic variant association study for 30 blood biomarkers in 191,971 individuals in the UK Biobank. We compare gene-based association tests for separate functional variant categories to increase interpretability and identify 193 significant gene-biomarker associations. Genes associated with biomarkers were ~ 4.5-fold enriched for conferring Mendelian disorders. In addition to performing weighted gene-based variant collapsing tests, we design and apply variant-category-specific kernel-based tests that integrate quantitative functional variant effect predictions for missense variants, splicing and the binding of RNA-binding proteins. For these tests, we present a computationally efficient combination of the likelihood-ratio and score tests that found 36% more associations than the score test alone while also controlling the type-1 error. Kernel-based tests identified 13% more associations than their gene-based collapsing counterparts and had advantages in the presence of gain of function missense variants. We introduce local collapsing by amino acid position for missense variants and use it to interpret associations and identify potential novel gain of function variants in PIEZO1. Our results show the benefits of investigating different functional mechanisms when performing rare-variant association tests, and demonstrate pervasive rare-variant contribution to biomarker variability.

Whole-genome sequencing identifies new candidate genes for nonobstructive azoospermia.

Genetic causes that lead to spermatogenetic failure in patients with nonobstructive azoospermia (NOA) have not been yet completely established. To identify low-frequency NOA-associated single nucleotide variants (SNVs) using whole-genome sequencing (WGS). Men with various types of NOA (n = 39), including samples that had been previously tested with whole-exome sequencing (WES; n = 6) and did not result in diagnostic conclusions. Variants were annotated using the Ensembl Variant Effect Predictor, utilizing frequencies from GnomAD and other databases to provide clinically relevant information (ClinVar), conservation scores (phyloP), and effect predictions (i.e., MutationTaster). Structural protein modeling was also performed. Using WGS, we revealed potential NOA-associated SNVs, such as: TKTL1, IGSF1, ZFPM2, VCX3A (novel disease causing variants), ESX1, TEX13A, TEX14, DNAH1, FANCM, QRICH2, FSIP2, USP9Y, PMFBP1, MEI1, PIWIL1, WDR66, ZFX, KCND1, KIAA1210, DHRSX, ZMYM3, FAM47C, FANCB, FAM50B (genes previously known to be associated with infertility) and ALG13, BEND2, BRWD3, DDX53, TAF4, FAM47B, FAM9B, FAM9C, MAGEB6, MAP3K15, RBMXL3, SSX3 and FMR1NB genes, which may be involved in spermatogenesis. In this study, we identified novel potential candidate NOA-associated genes in 29 individuals out of 39 azoospermic males. Note that in 5 out of 6 patients subjected previously to WES analysis, which did not disclose potentially causative variants, the WGS analysis was successful with NOA-associated gene findings.

ExAgBov: A public database of annotated variations from hundreds of bovine whole-exome sequencing samples

Large reference datasets of annotated genetic variations from genome-scale sequencing are essential for interpreting identified variants, their functional impact, and their possible contribution to diseases and traits. However, to date, no such database of annotated variation from broad cattle populations is publicly available. To overcome this gap and advance bovine NGS-driven variant discovery and interpretation, we obtained and analyzed raw data deposited in the SRA public repository. Short reads from 262 whole-exome sequencing samples of Bos Taurus were mapped to the Bos Taurus ARS-UCD1.2 reference genome. The GATK best practice workflow was applied for variant calling. Comprehensive annotation of all recorded variants was done using the Ensembl Variant Effect Predictor (VEP). An in-depth analysis of the population structure revealed the breeds comprising the database. The Exomes Aggregate of Bovine- ExAgBov is a comprehensively annotated dataset of more than 20 million short variants, of which ~2% are located within open reading frames, splice regions, and UTRs, and more than 60,000 variants are predicted to be deleterious.

Predicting the mutagenic effects of variants rs7188856 and rs7110 of N-Acetylglucosamine-1-Phosphodiester Alpha-N-Acetylglucosaminidase (NAGPA) gene detected in randomly collected blood samples

refore, though final confirmation of variant-effect predictions has to come from experiments and computational methods suggest what experiments to be performed

Candidate genes for tick resistance in cattle: a systematic review combining post-GWAS analyses with sequencing data

ABSTRACT Rhipicephalus microplus causes huge losses in cattle. Host genetic background greatly affects the immune efficiency in resistance or susceptibility to tick infestation, which is one of the many factors that play a role on that trait. We performed a systematic review of genome-wide association studies (GWAS) for tick resistance in cattle resulting in 1353 candidate genes for post-GWAS analyses. From those, genes showing possible structural variants from the bovine genome were classified by the Variant Effect Predictor from Ensembl. Ninety-two candidate genes showed potential structural variants in 5′ UTR and coding region and were used for functional annotation. Enriched biological processes (e.g. regulation of eosinophil chemotaxis, RIG-I signalling pathway and monocyte differentiation) and candidate genes (e.g. DAPK2, PUM1, ACIN1, INPP5D) linked with immune system function were identified and thus associated with tick resistance. Besides, gene-transcription factors (TFs) networks were obtained from TFs associated with immune system (FOXO3, PPARG, STAT3, NFKB1, GATA3 and ARNT) and the candidate genes associated with tick resistance in cattle highlighted (e.g. OR4L1, PNP, LRRIQ1, GIMAP8, MYO6, MEP1A and LRFN2). Thus, promising candidate genes with a possible functional role for tick resistance in cattle are presented for further in vitro and/or in vivo analyses.

Front Cover, Volume 43, Issue 8

Front Cover: The cover image is based on the Editorial Introduction Human mutation special issue on “variant effect prediction” by Ales Maver et al., https://doi.org/10.1002/humu.24429.

Loss-of-function, gain-of-function and dominant-negative mutations have profoundly different effects on protein structure

Most known pathogenic mutations occur in protein-coding regions of DNA and change the way proteins are made. Taking protein structure into account has therefore provided great insight into the molecular mechanisms underlying human genetic disease. While there has been much focus on how mutations can disrupt protein structure and thus cause a loss of function (LOF), alternative mechanisms, specifically dominant-negative (DN) and gain-of-function (GOF) effects, are less understood. Here, we investigate the protein-level effects of pathogenic missense mutations associated with different molecular mechanisms. We observe striking differences between recessive vs dominant, and LOF vs non-LOF mutations, with dominant, non-LOF disease mutations having much milder effects on protein structure, and DN mutations being highly enriched at protein interfaces. We also find that nearly all computational variant effect predictors, even those based solely on sequence conservation, underperform on non-LOF mutations. However, we do show that non-LOF mutations could potentially be identified by their tendency to cluster in three-dimensional space. Overall, our work suggests that many pathogenic mutations that act via DN and GOF mechanisms are likely being missed by current variant prioritisation strategies, but that there is considerable scope to improve computational predictions through consideration of molecular disease mechanisms.

Mutational load and deleterious mutations in goat genome from Kenya

Different domestication processes such as selection and effective population size are among other factors responsible for the accumulation of mutations in an animal. This study investigated the mutation load and accumulation of deleterious mutations in Kenyan goat populations using Single Nucleotide Polymorphism (SNP) data obtained from local (Galla = 12) and exotic (Saanen = 24, Alpine = 28, and Toggenburg = 30) goat genotypes. The SNP annotation was done on the ENSEMBLE goat (Capra hircus) using the Variant Effect Predictor (VEP). Sorting Intolerant from Tolerant (SIFT) was done on the annotated file using a score of > 0.01 to describe highly deleterious mutations. Biomart tool in the ENSEMBL goat C. hircus was used for Gene Ontology (GO) for the genes identified from highly deleterious mutations. The analysis disclosed no differences in mutation load among the studied genotypes. Overall, the synonymous mutations were in abundance compared to missense mutations, totaling 693 and 258 representing 1.01% and 0.37%, respectively. The calculated mutation load was similar (0.37) suggesting exposure of goats to similar domestication processes that are creating similar effects on the animal’s genome. Further analysis of the missense variants revealed 126 deleterious mutations and 132 tolerated mutations. A total of 111 genes were identified from the deleterious mutations and only 5 were among the highly deleterious mutations (SIFT score > 0.01) which include PROS1, EHBP1, LTN1, LRRN4, and FNDC3A. Gene Ontology describes these genes as responsible for different functions associated with animal reproduction, diseases, and memory. This study's results have confirmed the levels of mutation load and the presence of deleterious mutations in Kenyan goats. This can be used as a base to predict the future rate of mutation load to ensure better implementation of conservation programs to avoid an accumulation of deleterious mutations.

Heterogeneous Distribution of Genetic Mutations in Myosin Binding Protein-C Paralogs.

Myosin binding protein-C (MyBP-C) is a sarcomeric protein which regulates the force of contraction in striated muscles. Mutations in the MYBPC family of genes, including slow skeletal (MYBPC1), fast skeletal (MYBPC2) and cardiac (MYBPC3), can result in cardiac and skeletal myopathies. Nonetheless, their evolutionary pattern, pathogenicity and impact on MyBP-C protein structure remain to be elucidated. Therefore, the present study aimed to systematically assess the evolutionarily conserved and epigenetic patterns of MYBPC family mutations. Leveraging a machine learning (ML) approach, the Genome Aggregation Database (gnomAD) provided variants in MYBPC1, MYBPC2, and MYBPC3 genes. This was followed by an analysis with Ensembl’s variant effect predictor (VEP), resulting in the identification of 8,618, 3,871, and 3,071 variants in MYBPC1, MYBPC2, and MYBPC3, respectively. Missense variants comprised 61%–66% of total variants in which the third nucleotide positions in the codons were highly altered. Arginine was the most mutated amino acid, important because most disease-causing mutations in MyBP-C proteins are arginine in origin. Domains C5 and C6 of MyBP-C were found to be hotspots for most mutations in the MyBP-C family of proteins. A high percentage of truncated mutations in cMyBP-C cause cardiomyopathies. Arginine and glutamate were the top hits in fMyBP-C and cMyBP-C, respectively, and tryptophan and tyrosine were the most common among the three paralogs changing to premature stop codons and causing protein truncations at the carboxyl terminus. A heterogeneous epigenetic pattern was identified among the three MYBP-C paralogs. Overall, it was shown that databases using computational approaches can facilitate diagnosis and drug discovery to treat muscle disorders caused by MYBPC mutations.

Abstract 1964: Next-generation sequencing (NGS) reveals higher mitochondrial DNA (mtDNA) diversity in plasma extracellular vesicles (EVs) than in other blood components and tumor tissue from colorectal cancer (CRC) patients

Abstract Introduction: EVs released from cells are nanosized vesicles (containing proteins and nucleic acids) that serve as a communication system with other cells or expel cellular waste. Variations in the mtDNA sequence can act as functional adaptors allowing tumor and immune cells to adjust the metabolic needs imposed by various tissue environments during cancer progression. Recent reports have demonstrated that the entire mitochondrial genome can be encapsulated and secreted in EVs. However, the biological attributes of secreted cell-free mtDNA still remain insufficiently understood, mostly because of the difficulty in isolating and characterizing the entire mitochondrial genomes. Here, we present the results from sequencing of mtDNA from various blood compartments, including EVs, and tumor tissue from CRC patients. Methods: Plasma EVs (median mode size 136.4 nm) were isolated by size exclusion chromatography. We compared the mtDNA in EVs with that in whole blood (WB), peripheral blood mononuclear cells (PBMCs), and formalin-fixed paraffin-embedded (FFPE) tumor samples from eight rectal cancer patients. Total DNA was isolated from the four tissue types before the mtDNA was enriched by applying two different PCR approaches with either multiple (for FFPE tumor) or only two (for EVs, WB, and PBMCs) primer sets prior to NGS analysis. WB- and fresh-frozen (FF) tumor mtDNA from eight colon cancer patients was sequenced as controls for the two PCR primer approaches. mtDNA sequence mapping and variant analyses were done using the revised Cambridge Reference Sequence, and HaploGrep2- and Variant Effect Predictor softwares. Results: EV mtDNA presented twice as many variants with significantly more heteroplasmy (mutant mtDNA copies mixed with wild-type copies) than mtDNA from WB and PBMCs. The proportion of EV mtDNA variants that were non-synonymous (i.e. estimated to affect the mitochondrial function) was significantly higher than in WB. FFPE tumors, requiring the multi-primer approach, showed a manyfold increase in detected mtDNA variants, which was not observed for FF tumor mtDNA amplified like WB mtDNA. However, several mtDNA variants detected in FFPE- and FF tumors were not retrieved in WB. Conclusion: We successfully isolated and sequenced the mtDNA cargo of plasma EVs for comparison with mtDNA in WB, PBMCs, and tumor tissue from CRC patients. The EV mtDNA exhibited higher diversity than in the other blood compartments, with a higher degree of heteroplasmy and more variants with impact on the mitochondrial function. The majority of additional mtDNA variants detected in FFPE tumors were possibly caused by technical biases (the tissue processing or mtDNA amplification procedure). Plasma EVs may be involved in the metabolic homeostasis of tumor and immune cells in CRC patients. Citation Format: Tonje Bjørnetrø, Paula A. Bousquet, Kathrine R. Redalen, Anne-Marie S. Trøseid, Torben Lüders, Christin Johansen, Anniken J. Fuglestad, Christian Kersten, Sebastian Meltzer, Anne H. Ree. Next-generation sequencing (NGS) reveals higher mitochondrial DNA (mtDNA) diversity in plasma extracellular vesicles (EVs) than in other blood components and tumor tissue from colorectal cancer (CRC) patients [abstract]. In: Proceedings of the American Association for Cancer Research Annual Meeting 2022; 2022 Apr 8-13. Philadelphia (PA): AACR; Cancer Res 2022;82(12_Suppl):Abstract nr 1964.

An Active Learning Framework Improves Tumor Variant Interpretation.

A novel machine learning approach predicts the impact of tumor mutations on cellular phenotypes, overcomes limited training data, minimizes costly functional validation, and advances efforts to implement cancer precision medicine.