Nagarajan Paramasivam

A comprehensive assessment of somatic mutation detection in cancer using whole-genome sequencing

As whole-genome sequencing for cancer genome analysis becomes a clinical tool, a full understanding of the variables affecting sequencing analysis output is required. Here using tumour-normal sample pairs from two different types of cancer, chronic lymphocytic leukaemia and medulloblastoma, we conduct a benchmarking exercise within the context of the International Cancer Genome Consortium. We compare sequencing methods, analysis pipelines and validation methods. We show that using PCR-free methods and increasing sequencing depth to ∼100 × shows benefits, as long as the tumour:control coverage ratio remains balanced. We observe widely varying mutation call rates and low concordance among analysis pipelines, reflecting the artefact-prone nature of the raw data and lack of standards for dealing with the artefacts. However, we show that, using the benchmark mutation set we have created, many issues are in fact easy to remedy and have an immediate positive impact on mutation detection accuracy.

The whole-genome landscape of medulloblastoma subtypes

Current therapies for medulloblastoma, a highly malignant childhood brain tumour, impose debilitating effects on the developing child, and highlight the need for molecularly targeted treatments with reduced toxicity. Previous studies have been unable to identify the full spectrum of driver genes and molecular processes that operate in medulloblastoma subgroups. Here we analyse the somatic landscape across 491 sequenced medulloblastoma samples and the molecular heterogeneity among 1,256 epigenetically analysed cases, and identify subgroup-specific driver alterations that include previously undiscovered actionable targets. Driver mutations were confidently assigned to most patients belonging to Group 3 and Group 4 medulloblastoma subgroups, greatly enhancing previous knowledge. New molecular subtypes were differentially enriched for specific driver events, including hotspot in-frame insertions that target KBTBD4 and ‘enhancer hijacking’ events that activate PRDM6. Thus, the application of integrative genomics to an extensive cohort of clinical samples derived from a single childhood cancer entity revealed a series of cancer genes and biologically relevant subtype diversity that represent attractive therapeutic targets for the treatment of patients with medulloblastoma.

Loss of function of PGAP1 as a cause of severe encephalopathy identified by Whole Exome Sequencing: Lessons of the bioinformatics pipeline.

We evaluated a multiple consanguineous Turkish family with two children, a boy and a girl, affected by severe encephalopathy, hypotonia, microcephaly and retinal dystrophy by a combination of linkage analysis and Whole Exome Sequencing (WES). We analyzed the sequence data by two different bioinformatics pipelines which did not differ in overall processing strategy but involved differences in software used, minor allele frequency (MAF) thresholds and reference data sets, the usage of in-house control exomes and filter settings to prioritize called variants. Assuming autosomal recessive mode of inheritance, only homozygous variants present in both children were considered. The resulting variant lists differed partially (nine variants identified by both pipelines, ten variants by only one pipeline). Major reasons for this discrepancy were different filters for MAF and different variant prioritizations. Combining the variant lists with the results of linkage analysis and further prioritization by expression data and prediction tools, an intronic homozygous splice variant (c.1090-2A>G; IVS9-2A>G; p.?) in PGAP1 (Post-GPI Attachment To Proteins 1) was identified and validated by cDNA analysis. PGAP1 ensures the first step of maturation of GPI (glycosylphosphatidylinositol)-anchor proteins. Recently, a homozygous loss-of-function mutation in PGAP1 has been reported in one family with two children affected by a similar phenotype. The present report not only illustrates the possible influence of specific filtering settings on the results of WES but also confirms PGAP1 as a cause of severe encephalopathy.

SIPA1L3 identified by linkage analysis and whole-exome sequencing as a novel gene for autosomal recessive congenital cataract

Congenital cataract (CC) is one of the most important causes for blindness or visual impairment in infancy. A substantial proportion of isolated CCs has monogenic causes. The disease is genetically heterogeneous, and all Mendelian modes of inheritance have been reported. We mapped a locus for isolated CC on 19p13.1–q13.2 in a distantly consanguineous German family with two sisters affected by dense white cataracts. Whole-exome sequencing identified a homozygous nonsense variant c.4489C>T (p.(R1497*)) in SIPA1L3 (signal-induced proliferation-associated 1 like 3) in both affected children. SIPA1L3 encodes a GTPase-activating protein (GAP), which interacts with small GTPases of the Rap family via its Rap-GAP-domain. The suggested role of Rap GTPases in cell growth, differentiation and organization of the cytoskeleton in the human lens, and lens-enriched expression of the murine ortholog gene Sipa1l3 in embryonic mice indicates that this gene is crucial for early lens development. Our results provide evidence that sequence variants in human SIPA1L3 cause autosomal recessive isolated CC and give new insight into the molecular pathogenesis underlying human cataracts.

Impact of clinical exomes in neurodevelopmental and neurometabolic disorders

Whole exome sequencing (WES) is well established in research and is now being introduced into clinically indicated diagnostics (so-called clinical exomes). We evaluated the diagnostic yield and clinical implications of WES in 72 patients from 60 families with undiagnosed neurodevelopmental disorders (NDD), neurometabolic disorders, and dystonias. Pathogenic or likely pathogenic variants leading to a molecular diagnosis could be identified in 21 of the 60 families (overall 35%, in 36% of patients with NDD, in 43% of patients with neurometabolic disorders, in 25% of patients with dystonias). In one family two coexisting autosomal recessive diseases caused by homozygous pathogenic variants in two different genes were diagnosed. In another family, a homozygous frameshift variant in STRADA was found to cause a severe NDD with early onset epilepsy, brain anomalies, hypotonia, heart defect, nephrocalcinosis, macrocephaly and distinctive facies so far designated as PMSE (polyhydramnios, megalencephaly, symptomatic epilepsy) syndrome. In 7 of the 21 families with a molecular diagnosis the pathogenic variants were only identified by clinical follow-up, manual reevaluation of the literature, a change of filter setting, and/or reconsideration of inheritance pattern. Most importantly, clinical implications included management changes in 8 cases and impact on family planning in 20 families with a molecular diagnosis. This study shows that reevaluation and follow-up can improve the diagnostic rate and that WES results have important implications on medical management and family planning. Furthermore, we could confirm STRADA as a gene associated with syndromic ID but find it questionable if the current designation as PMSE depicts the most important clinical features.

DDX3X mutations in two girls with a phenotype overlapping Toriello–Carey syndrome

Recently, de novo heterozygous variants in DDX3X have been reported in about 1.5% of 2659 females with previously unexplained intellectual disability (ID). We report on the identification of DDX3X variants in two unrelated girls with clinical features of Toriello–Carey Syndrome (T‐CS). In patient 1, the recurrent variant c.1703C>T; p.(P568L) was identified when reconsidering X‐linked de novo heterozygous variants in exome sequencing data. In patient 2, the DDX3X variant c.1600C>G; p.(R534G) was also detected by exome sequencing. Based on these data, de novo heterozygous DDX3X variants should be considered not only in females with unexplained ID, but also in individuals with a clinical diagnosis of T‐CS.

Diagnosis of CoPAN by whole exome sequencing: Waking up a sleeping tiger's eye

Neurodegeneration with brain iron accumulation (NBIA) is a group of neurodegenerative disorders characterized by iron accumulation in the basal ganglia. Recently, mutations in CoA synthase (COASY) have been identified as a cause of a novel NBIA subtype (COASY Protein‐Associated Neurodegeneration, CoPAN) in two patients with dystonic paraparesis, parkinsonian features, cognitive impairment, behavior abnormalities, and axonal neuropathy. COASY encodes an enzyme required for Coenzyme A (CoA) biosynthesis. Using whole exome sequencing (WES) we identified compound heterozygous COASY mutations in two siblings with intellectual disability, ataxic gait, progressive spasticity, and obsessive‐compulsive behavior. The “eye‐of‐the tiger‐sign,” a characteristic hypointense spot within the hyperintense globi pallidi on MRI found in the most common subtype of NBIA (Pantothenate Kinase‐Associated Neurodegeneration, PKAN), was not present. Instead, bilateral hyperintensity and swelling of caudate nucleus, putamen, and thalamus were found. In addition, our patients showed a small corpus callosum and frontotemporal and parietal white matter changes, expanding the brain phenotype of patients with CoPAN. Metabolic investigations showed increased free carnitine and decreased acylcarnitines in the patientś dried blood samples. Carnitine palmitoyl transferase 1 (CPT1) deficiency was excluded by further enzymatic and metabolic investigations. As CoA and its derivate Acetyl‐CoA play an essential role in fatty acid metabolism, we assume that abnormal acylcarnitine profiles are a result of the COASY mutations. This report not only illustrates that WES is a powerful tool to elucidate the etiology of rare genetic diseases, but also identifies unique neuroimaging and metabolic findings that may be key features for an early diagnosis of CoPAN.

Homozygous missense mutation in the LMAN2L gene segregates with intellectual disability in a large consanguineous Pakistani family

Background Intellectual disability (ID) is a neurodevelopmental disorder affecting 1%–3% of the population worldwide. It is characterised by high phenotypic and genetic heterogeneity and in most cases the underlying cause of the disorder is unknown. In our study we investigated a large consanguineous family from Baluchistan, Pakistan, comprising seven affected individuals with a severe form of autosomal recessive ID (ARID) and epilepsy, to elucidate a putative genetic cause. Methods and results Whole exome sequencing (WES) of a trio, including a child with ID and epilepsy and its healthy parents that were part of this large family, revealed a homozygous missense variant p.R53Q in the lectin mannose-binding 2-like (LMAN2L) gene. This homozygous variant was co-segregating in the family with the phenotype of severe ID and infantile epilepsy; unaffected family members were heterozygous variant carriers. The variant was predicted to be pathogenic by five different in silico programmes and further three-dimensional structure modelling of the protein suggests that variant p.R53Q may impair protein–protein interaction. LMAN2L (OMIM: 609552) encodes for the lectin, mannose-binding 2-like protein which is a cargo receptor in the endoplasmic reticulum important for glycoprotein transport. Genome-wide association studies have identified an association of LMAN2L to different neuropsychiatric disorders. Conclusion This is the first report linking LMAN2L to a phenotype of severe ARID and seizures, indicating that the deleterious homozygous p.R53Q variant very likely causes the disorder.

A Comprehensive Assessment of Somatic Mutation Calling in Cancer Genomes

The emergence of next generation DNA sequencing technology is enabling high-resolution cancer genome analysis. Large-scale projects like the International Cancer Genome Consortium (ICGC) are systematically scanning cancer genomes to identify recurrent somatic mutations. Second generation DNA sequencing, however, is still an evolving technology and procedures, both experimental and analytical, are constantly changing. Thus the research community is still defining a set of best practices for cancer genome data analysis, with no single protocol emerging to fulfil this role. Here we describe an extensive benchmark exercise to identify and resolve issues of somatic mutation calling. Whole genome sequence datasets comprising tumor-normal pairs from two different types of cancer, chronic lymphocytic leukaemia and medulloblastoma, were shared within the ICGC and submissions of somatic mutation calls were compared to verified mutations and to each other. Varying strategies to call mutations, incomplete awareness of sources of artefacts, and even lack of agreement on what constitutes an artefact or real mutation manifested in widely varying mutation call rates and somewhat low concordance among submissions. We conclude that somatic mutation calling remains an unsolved problem. However, we have identified many issues that are easy to remedy that are presented here. Our study highlights critical issues that need to be addressed before this valuable technology can be routinely used to inform clinical decision-making. Abbreviations and Definitions SSM Somatic Single-base Mutations or Simple Somatic Mutations, refers to a somatic single base change SIM Somatic Insertion/deletion Mutation CNV Copy Number Variant SV Structural Variant SNP Single Nucleotide Polymorphisms, refers to a single base variable position in the germline with a frequency of > 1% in the general population CLL Chronic Lymphocytic Leukaemia MB Medulloblastoma ICGC International Cancer Genome Consortium BM Benchmark aligner = mapper, these terms are used interchangeably

Genetic subclone architecture of tumor clone-initiating cells in colorectal cancer

A hierarchically organized cell compartment drives colorectal cancer (CRC) progression. Genetic barcoding allows monitoring of the clonal output of tumorigenic cells without prospective isolation. In this study, we asked whether tumor clone-initiating cells (TcICs) were genetically heterogeneous and whether differences in self-renewal and activation reflected differential kinetics among individual subclones or functional hierarchies within subclones. Monitoring genomic subclone kinetics in three patient tumors and corresponding serial xenografts and spheroids by high-coverage whole-genome sequencing, clustering of genetic aberrations, subclone combinatorics, and mutational signature analysis revealed at least two to four genetic subclones per sample. Long-term growth in serial xenografts and spheroids was driven by multiple genomic subclones with profoundly differing growth dynamics and hence different quantitative contributions over time. Strikingly, genetic barcoding demonstrated stable functional heterogeneity of CRC TcICs during serial xenografting despite near-complete changes in genomic subclone contribution. This demonstrates that functional heterogeneity is, at least frequently, present within genomic subclones and independent of mutational subclone differences.