Araujo P.A.T.,Paediatric Laboratory Medicine |
Thomas D.,Paediatric Laboratory Medicine |
Sadeghieh T.,Paediatric Laboratory Medicine |
Bevilacqua V.,Paediatric Laboratory Medicine |
And 4 more authors.
Clinical Biochemistry | Year: 2015
Background: The CALIPER program has established a comprehensive database of age- and sex-stratified pediatric reference intervals for over 85 common biochemical markers, largely using the Abbott ARCHITECT assays. To allow a broader application of the CALIPER database, we examined transference to 36 Beckman Coulter Synchron Unicel DxC800 assays, based on the CLSI C28-A3/EP9-A3 guidelines. Methods: Patient sample comparisons were performed for 36 biochemical assays using 200 serum specimens obtained from pediatric patients on the Abbott ARCHITECT ci8200 and the Beckman Coulter DxC800. For each analyte, R2 values were calculated to assess the quality of correlation between the platforms. Statistical criteria used to assess transferability included a) regression analysis to create the equation of the line of best fit, b) standardized residual, c) Bland-Altman, and d) quantile-quantile plots. Transferred reference intervals were further verified by analyzing serum samples from 100 healthy children from the CALIPER cohort on the Beckman Coulter system. Results: The reference intervals for most of the assessed analytes were transferable to Beckman Coulter assays (31 out of 36 studied) and the newly calculated reference intervals were verified through analysis of CALIPER reference samples (28 out of 31). Eighteen assays demonstrated excellent correlation (R2≥0.95), and 13 assays showed strong correlation (0.77≤R2≤0.94). Conclusion: The current study allowed successful transference of a large number of biochemical markers from the CALIPER database to assays on the Beckman Coulter DxC800 platform. Transference should facilitate broader application of CALIPER reference intervals at pediatric centers using DxC biochemical assays. © 2015 The Canadian Society of Clinical Chemists.
Addis L.,Kings College London |
Addis L.,Eli Lilly and Company |
Ahn J.W.,Guys and St Thomas NHS Foundation Trust |
Dobson R.,Kings College London |
And 23 more authors.
Human Mutation | Year: 2015
Copy-number variations (CNVs) are important in the aetiology of neurodevelopmental disorders and show broad phenotypic manifestations. We compared the presence of small CNVs disrupting the ELP4-PAX6 locus in 4,092 UK individuals with a range of neurodevelopmental conditions, clinically referred for array comparative genomic hybridization, with WTCCC controls (n = 4,783). The phenotypic analysis was then extended using the DECIPHER database. We followed up association using an autism patient cohort (n = 3,143) compared with six additional control groups (n = 6,469). In the clinical discovery series, we identified eight cases with ELP4 deletions, and one with a partial duplication of ELP4 and PAX6. These cases were referred for neurological phenotypes including language impairment, developmental delay, autism, and epilepsy. Six further cases with a primary diagnosis of autism spectrum disorder (ASD) and similar secondary phenotypes were identified with ELP4 deletions, as well as another six (out of nine) with neurodevelopmental phenotypes from DECIPHER. CNVs at ELP4 were only present in 1/11,252 controls. We found a significant excess of CNVs in discovery cases compared with controls, P = 7.5 × 10-3, as well as for autism, P = 2.7 × 10-3. Our results suggest that ELP4 deletions are highly likely to be pathogenic, predisposing to a range of neurodevelopmental phenotypes from ASD to language impairment and epilepsy. We have identified a significant excess (p = 7.5 × 10-3) of small deletions (shown as red lines on the figure) at the PAX6-ELP4 locus, 11p13, in three cohorts of patients with neurodevelopmental disorders. The deletions predispose to a range of phenotypes including autism spectrum disorder, language impairment, mental retardation and epilepsy, and likely disrupt the functions of the Elongator protein complex and/or the transcription factor PAX6. © 2015 WILEY PERIODICALS, INC..
Adam H.J.,Health Science Center |
Richardson S.E.,Public Health England |
Richardson S.E.,University of Toronto |
Richardson S.E.,Paediatric Laboratory Medicine |
And 7 more authors.
Vaccine | Year: 2010
The epidemiology of invasive Haemophilus influenzae infections was evaluated in Ontario between 1989 and 2007 to assess the impact of the introduction of the conjugate H. influenzae serotype b (Hib) vaccine in the early 1990s on Hib and non-Hib serotypes in both vaccinated and unvaccinated cohorts as well as the possibility of "strain replacement" with non-vaccine H. influenzae strains. Data were collected by the provincial Public Health Laboratories-Toronto, Ontario Agency for Health Protection and Promotion, which performed almost all serotyping on invasive (blood, CSF, other sterile sites) H. influenzae strains isolated in the province during the study period. Temporal trends for Hib, other typeable strains, and non-typeable H. influenzae were evaluated by Poisson regression, controlling for the specimen submissions. Prior to infant Hib vaccination, the most commonly observed serotype was serotype b (64.9%). Subsequently, 70.3%, 13.6%, and 9.4% of isolates were non-typeable, serotype f, and serotype b, respectively. Infant Hib vaccination resulted in a decrease in Hib incidence in all age groups (pooled IRR 0.432) and marked increases of non-typeable and serotype f H. influenzae in children aged <5 years (IRR 2.4 and 3.0, respectively). Vaccination against Hib has altered the epidemiology of invasive H. influenzae infections in Ontario. Prevention of invasive Hib disease was observed in both vaccinated and unvaccinated age groups. Invasive H. influenzae infection now commonly presents as sepsis due to non-typeable H. influenzae in older individuals. However, strain replacement of Hib with serotype f and non-typeable strains in children under 5 years was documented. © 2010 Elsevier Ltd. All rights reserved.
No statistical methods were used to predetermine sample size. Male and female Ptc+/−/Math1-SB11/T2Onc or T2Onc2 mice (12 to 20 weeks of age; at the time they developed signs of medulloblastoma) were used. We did not perform a formal sample size estimate for the study but based our experimental plan on our previous experience with Sleeping Beauty mutagenesis screening. When mice showed early clinical signs of brain tumours they were anaesthetized with isoflourane, ophthalmic ointment applied to the eyes and the scalp antiseptically prepared. A 1.5 cm long midline incision was made to expose the skull from the coronal suture to the cranio-cervical junction. Using a high-speed drill and a 2.5 mm trephine bit, a cranial defect is drilled 2 mm posterior to lambda to avoid the transverse sinuses. The skull and the dura are lifted with micro-dissecting forceps, the bulk of the tumour is then removed using a harmon forceps with teeth, while smaller sections of tumour are removed with a microcurette (2 mm). Surgical samples are saved in dry ice, the bleeding from the tumour site is counteracted with direct pressure and Gelfoam. When haemostasis is obtained, the surgical wound is sutured using interrupted stitching with absorbable sutures. Animals received analgesia and dexamethasone post-operatively to contain the brain oedema. Male and female Ptc+/−/Math1-SB11/T2Onc or T2Onc2 control mice were monitored for early clinical signs of brain tumours but not subjected to surgery and CSI irradiation, no formal randomization was used. All the procedures involving animals have been approved by the institutional Animal Care Committee, in no case were tumour-bearing animals allowed a tumour burden compromising normal behaviour, food and water intake or exceeding the approved volume of 1,700 mm3. Mice that had recovered from tumour resection were anaesthetized with isoflurane and placed in the brain irradiation bed in the image guided small animal irradiator (X-Rad 225CX, Precision Xray, North Branford, CT, USA). Correct animal setup was confirmed using 2D fluoroscopic images with and without the brain collimator (2 × 2 cm) in place, all images were acquired at 40 kVp, 0.5 mA, using the same X-ray tube which is used for radiation treatment. 3D volumetric cone-beam CT images were used for the visualization of bone and soft tissues within the animal and isocentre placement. The imaging capability of the unit were described previously43, the imaging dose to the animal was estimated to be less than 1 cGy. The delivered dose per fraction was 2 Gy, administered 3 times a week for the first week to prevent brain oedema, followed by five times a week treatment for the following 3 weeks. Each daily dose was delivered with two parallel opposed-lateral beams to correct for tissue attenuation of the dose, total daily dose of 2 Gy. Dose rate for the brain collimator was measured at 3.2 Gy per min at 225 kVp, 13 mA, on a 0.3 mm Cu filter (HVL: 0.93 mm Cu, added filtration: 0.3 mm Cu). The tube was calibrated at these settings following the TG61 protocol44. The spine treatment was introduced on the second week of CSI irradiation, we used a 4.76Gy per 6 fractions schedule, and the mice received 2 spinal fractions per week. Radiation to the spinal cord was delivered to mice placed supine on the irradiator stage the irradiation was done with single or multiple posterior beams. The same imaging strategy with 2D and volumetric 3D imaging was adopted for spinal cord targeting, using a 0.5 × 5 cm collimator or multiple fields of 0.5 × 2 cm; for the spine treatment a dose correction was applied to compensate for the different depth of the cervical spine compared to lumbo/sacral. Treatment dose was administered at 2.8Gy per min at 225 kVp, 13 mA settings on a 0.3 mm Cu filter. The end-point date of the control and CSI treated groups was assessed by independent veterinary technicians blinded to the experimental group. Medulloblastoma-free survival from the time of diagnosis was assessed for control mice and mice that underwent surgery and radiation, no animal was removed from the study and mice euthanized during the study for different reasons than medulloblastoma were censored in the Kaplan–Meier estimate for tumour-free survival. Genomic DNA was isolated and purified from mouse tissues with a PureLink genomic DNA extraction kit (Invitrogen). Libraries for Illumina HiSeq sequencing were prepared as described previously25 with minor modifications. 2 μg of gDNA were digested and ligated to the adapters, after a BamHI digest to eliminate the untransposed copies of the concatamer, an enrichment PCR followed by a barcoding PCR were performed25. The barcode PCR was modified to incorporate a paired-end (PE) sequencing adaptor for paired end sequencing, the sequence of the PE adaptor was: CAAGCAGAAGACGGCATACGAGATCGGTCTCGGCATTCCTGCTGAACCGCTCTTCCGATCTTAGGGCTCCGCTTAAGGGAC. Libraries were purified and pooled as previously described and sequenced on an Illumina HiSeq 2000 (ref. 25). Sequenced libraries were demultiplexed and aligned as described previously25. Demultiplexing and trimming of SB transpon sequences was performed using custom scripts, alignment of reads was performed with Novoalign to mouse assembly NCBI37/mm9 (July 2007) A chi-squared test was used to asses statistical enrichment of the integration events within each transcription unit considering the following: the number of TA dinucleotide sites within the gene relative to the number of TA sites in the genome, the number of integration sites within each tumour, and the total number of tumours in each cohort. This gCIS analysis produced a P value for each of the 19,000 mouse RefSeq genes, and a Bonferroni correction was therefore used to adjust for multiple hypothesis testing. gCIS predictions were manually curated to filter out ambiguities, artefacts and local hopping. BioProject ID PRJNA306269. The model assumes that tumour cell division and growth are initiated by a founding transposon insertion event, and that additional insertion events can subsequently occur in daughter cells. According to the model, insertion events in the transformed daughter cells are expected to decrease by a factor of 2n relative to the initial transformed cell, where n is the number of intervening cell divisions. Details of the model are described in Supplementary Note 1. As with any model it is important to note its limitations. First, there is a limit to the degree to which distinct lineages can be separated. If two lineages are governed by two sufficiently close values of the parameter G, the components will be superimposed. If the value of d is also the same, the identification of the initiating insertions will not be affected; otherwise, the lineage with lower d will incorrectly identify its initiating mutation as a passenger. The extent of this issue is dependent upon the closeness of G and the depth at which a sample has been sequenced. It almost certainly true that other lineages are present in the data, but arose relatively late and/or have relatively low growth rates. Therefore, the model is best described as identifying the most clear and unambiguous lineages. Second, a lineage which have undergone multiple gene disruptions that affect growth rate at different generations can appear as two separate lineages. For example, if a disruption of gene A causes rampant cell division/growth, and is followed up two generations later by a disruption in gene B that further increases the growth rate, this will appear as two lineages with putative genotypes A- and B-. In reality, the genotypes are A- and A-;B-. Importantly, this does not affect the ultimate identification of both of these genes as initiators. Relative clonal prevalence was calculated for the genes predicted as driver as:2d G and normalized to the total number of predicted drivers for each sample. Driver events predicted to happen in the founder clones (highest G) for each sample, or showing relative cell abundance >10% were selected for pathway enrichment analysis. The primers for amplifying Sleeping Beauty transposon insertion sites were designed based on the chromosomal location of each insertion site and its orientation to the transcription of the gene hosting the insertion. The primers at the inverted repeats/direct repeats (left) (IRL) and inverted repeats/direct repeats (right) (IRR) of the transposon were 5′-AAATTTGTGGAGTAGTTGAAAAACGA-3′ and 5′-GGATTAAATGTCAGGAATTGTGAAAA-3′, respectively. The input represents genomic DNA with Sleeping Beauty transposition, which was illustrated by Sleeping Beauty excision PCR that detected the transposon post-transposition26. Three points of input (1×, 5× and 25×) were used. Mice showing signs of late stage brain tumours were euthanized and tissue harvested for genomic DNA extraction as well as histological examination. Extent and location of recurrences was evaluated by standard haematoxylin and eosin staining, Trp53 pathway status was evaluated by p21 staining performed at the Paediatric Laboratory Medicine Department, The Hospital for Sick Children, (Toronto, Canada) using the Ventana BenchMark XT model. The conditions were as follows: HIER: 40 min in a Tris based buffer (pH 8.5) Ventana CC1 (http://www.ventana.com/product/203?type=204), primary antibody p21 (1:50) (BD bioscience 556431, clone SXM30) was incubated for 1 h at 37 °C. The signal was detected using Ventana OptiView DAB IHC Detection Kit. The following fly stocks were used: UAS-mCD8-GFP to label cell membrane; insc-Gal4 (Gal41407 inserted in an inscuteable promoter) to drive gene expression in the neuroblast lineage; UAS-dpn for overexpression of dpn. Flies were mated and maintained at 25 °C. Fly larvae were retrieved at late third instar stage for whole body irradiation at 40 Gy. The larval brains were dissected 4 h after irradiation, followed by fixation and immunohistochemistry analysis. Larval brains were dissected, fixed, and stained as previously described29. Briefly, third instar larvae brains were dissected in PBS, fixed in 4% paraformaldehyde solution for 20 min at room temperature, and incubated with the primary antibody (rabbit anti-phospho-histone 3, Millipore, 1:200) overnight at 4 °C and secondary antibody for 2 h at room temperature. Fluorescence images were acquired using a Leica SP5 confocal microscope. Representative images of the dorsal brain lobes were shown in Fig. 1d, e and Extended Data Fig. 1h. All patients gave informed consent to the samples collection; unless indicated otherwise, the samples were sequenced and analysed at Canada’s Michael Smith Genome Sciences Centre at the BC Cancer Agency (GSC). Libraries for whole-genome sequencing were constructed using either the plate-based or SPRI-TE library construction protocol. 2 μg of genomic DNA in a 96-well format was fragmented by Covaris E210 sonication for 30 s using a ‘duty cycle’ of 20% and ‘intensity’ of 5. The paired-end sequencing library was prepared following the BC Cancer Agency’s Genome Sciences Centre 96-well genomic ~350–450 bp insert Illumina Library Construction protocol on a Biomek FX robot (Beckman-Coulter, USA). Briefly, the DNA was purified in a 96-well microtitre plate using Ampure XP SPRI beads (40–45 μl beads per 60 μl DNA), and was subject to end-repair, and phosphorylation by T4 DNA polymerase, Klenow DNA Polymerase, and T4 polynucleotide kinase respectively in a single reaction, followed by cleanup using Ampure XP SPRI beads and 3′ A-tailing by Klenow fragment (3′ to 5′ exo minus). After cleanup using Ampure XP SPRI beads, PicoGreen quantification was performed to determine the amount of Illumina PE adapters used in the next step of adaptor ligation reaction. The adaptor-ligated products were purified using Ampure XP SPRI beads, then PCR-amplified with Phusion DNA Polymerase (Thermo Fisher Scientific Inc. USA) using Illumina’s PE indexed primer set, with cycle conditions: 98 °C for 30 s followed by 6 cycles of 98 °C for 15 s, 62 °C for 30 s and 72 °C for 30 s, and a final extension at 72 °C for 5 min. The PCR products were purified using Ampure XP SPRI beads, and checked with Caliper LabChip GX for DNA samples using the High Sensitivity Assay (PerkinElmer, USA). PCR products of the desired size range were gel purified (8% PAGE or 1.5% Metaphor agarose in an in-house custom built robot), and the DNA quality was assessed and quantified using an Agilent DNA 1000 series II assay and Quant-iT dsDNA HS Assay Kit using Qubit fluorometer (Invitrogen), then diluted to 8 nM. The final concentration was confirmed by Quant-iT dsDNA HS Assay before generating 100 bp paired-end reads on the Illumina HiSeq 2000/2500 platform using v3 chemistry. Whole-genome libraries of patient samples medulloblastoma-Rec-03, -04, -06, -11, -12, -18, -19, -22–24, -26-33 have been constructed using the Spri-TE 300-600 bp fragment protocol as follows. Genome libraries with fragment size ranges of approximately 400 bp were constructed on a SPRI-TE robot (Beckman Coulter, USA) according to the manufacturer’s instructions (SPRIworks Fragment Library System I Kit, A84801). Briefly, 1 μg of genomic DNA in a 60 μl volume, and 96-well format, was fragmented by Covaris E210 sonication for 30 s using a ‘duty cycle’ of 20% and ‘intensity’ of 5. Up to 10 paired-end genome sequencing libraries were prepared in parallel using the SPRI-TE 300–600 bp size-selection program. Following completion of the SPRI-TE run the adaptor ligated library templates were quantified using a Qubit fluorometer. 5 ng of adaptor ligated template was PCR amplified using Phusion DNA Polymerase (Thermo Fisher Scientific, USA) and Illumina’s PE indexed primer set, with cycle conditions: 98 °C for 30 s followed by 10 cycles of 98 °C for 15 s, 62 °C for 30 s and 72 °C for 30 s, and a final amplicon extension at 72 °C for 5 min. The PCR products were purified using Ampure XP SPRI beads, and checked with Caliper LabChip GX for DNA samples using the High Sensitivity Assay (PerkinElmer, USA). PCR products of the desired size range were purified using gel electrophoresis (8% PAGE or 1.5% Metaphor agarose gels in a custom built robot) and the DNA quality was assessed and quantified using an Agilent DNA 1000 series II assay and Quant-iT dsDNA HS Assay Kit using Qubit fluorometer (Invitrogen), then diluted to 8 nM. The final concentration was verified by Quant-iT dsDNA HS Assay before Illumina Sequencing before generating 100 bp paired-end reads on the Illumina HiSeq 2000/2500 platform using v2 or v3 chemistry. Alignment. After marking chastity failed reads, paired-end 100 bp raw reads were aligned to the reference genome GRCh37-lite (http://www.bcgsc.ca/downloads/genomes/9606/hg19/1000genomes/bwa_ind/genome) with the Burrows–Wheeler Aligner (BWA; version 0.5.7)45. Bam files were sorted with SAMTools (version 0.1.13) and merged using Picard MarkDuplicates.jar (version 1.71). The merged bam files were subsequently indexed with SAMTools index (version 0.1.17) and submitted to the European Genome-phenome Archive (EGAD00001000946). German Cancer Research Centre (DKFZ). Patient samples medulloblastoma-REC-13-16 and medulloblastoma-REC-34-35 were processed at the DKFZ in Heidelberg as previously described2. Analysed DNA was isolated using using a Qiagen Allprep DNA/RNA/Protein Mini Kit. On average 125 mg of homogenized (TissueLyser, Qiagen) tumour tissue was used for isolation of analytes. The manufacturer’s protocol was adapted to allow for DNA and total RNA (including miRNA) isolation. DNA from matching blood samples was extracted using Qiagen Blood and Cell Culture Midi Kit according to the manufacturer’s protocol. After quality control of isolated DNA (gel electrophoresis), extracted nucleic acids were submitted for sequencing. Paired-end (PE) DNA library preparation was carried out using Illumina Inc. v2 protocols. In brief, 1-5μg of genomic DNA were fragmented to ~300 bp (PE) insert-size with a Covaris device, followed by size selection through agarose gel excision. Deep sequencing was carried out with Illumina HiSeq2000 instruments. Whole-exome sequencing at McGill. Patient samples medulloblastoma-REC-36-38 and medulloblastoma-REC-48-55 were prepared and sequenced by the Genome Quebec Innovation Centre and analysed at the McGill University Health Centre as follows. Paired-end libraries were prepared with the Illumina’s Nextera Rapid Capture Exome kit. Captured exome DNA fragments were then sequenced on Illumina HiSeq 2500 (rapid-run mode) generating 100-bp paired-end reads. Adaptor sequences were removed and low-quality reads were trimmed using the FASTX toolkit. Quality trimmed reads were aligned to the human genome reference library (hg19) using Burrows–Wheeler Aligner (BWA) version 0.5.9 (ref. 45). Indels were realigned using the Genome Analysis Toolkit (GATK)46 and duplicate reads were marked using Picard. SNVs from WGS data were analysed using all three methods described below, whereas SNVs from exome-seq data were analysed only with MutationSeq. SNVs were analysed with SAMtools mpileup v.0.1.17 either on single or paired libraries. Each chromosome was analysed separately using the -C50-DSBuf parameters. The resulting vcf files were merged and filtered to remove low-quality SNVs by using samtools varFilter (with default parameters) as well as to remove SNVs with a QUAL score of less than 20. Finally, SNVs were annotated with gene annotations from Ensembl v66 using snpEff and the dbSNP v137 db membership assigned using SnpSift47. To analyse compartment specific SNVs and indels, samples were analysed pair-wise with the default settings of Strelka v0.4.7 (ref. 48). Primary tumour samples and relapse/met were compared against the germline sample. In the absence of a germline sample, the relapse/met samples were compared against the primary tumour sample. Variant allele frequencies (VAF) of somatic damaging SNVs (called by Strelka in 14 patients with matched germline samples) were classified into distinct clusters using the R package mclust, which uses finite mixture estimation via iterative expectation maximization steps (EM) and the Bayesian Information Criterion (BIC). Each cluster is manually categorized as either ‘homozygous’, ‘clonal’, or ‘subclonal’, depending on the cluster VAF and the uncertainty separating it from the next cluster. Multiple subclonal populations are numbered sequentially, starting with the most highly prevalent population. SNVs were analysed pair-wise with SAMtools mpileup v.0.1.17 (ref. 49). Each chromosome was analysed separately using the -C50-DSBuf parameters. Before merging the resulting vcf files, they were filtered to remove all indels and low quality SNVs by using samtools varFilter (with default parameters) as well as to remove SNVs with a QUAL score of less than 20 (vcf column 6). The SNVs in the resulting vcf files were further filtered and scored using mutationSeq v1.0.2 and annotated with gene annotations from Ensembl v66 using SnpEff and the dbSNP v137 and Cosmic 64 db membership using SnpSift Indels were called in the low quality exomes using VarScan version 2.3.6, using the following parameters: P value 95 × 10−2 –strand-filter 1–min-avg-qual 20. The indels in the resulting vcf files were annotated with gene annotations from Ensembl 66 using SnpEff as described above, and screened against dbSNP137 using SnpSift. EMu was used to define mutation spectra for 11 samples with germline (that is, excluding the DKFZ samples), using the expectation-maximization algorithm50. To assess significant changes in the distributions of mutation spectra across primary, local and distal recurrences from each medulloblastoma patient, we used the chi-squared test. Changes in (1) the number of compartment-specific mutations and (2) in frequencies of transversion mutations, were tested with the Wilcoxon rank-sum test. Changes in the frequency of C > T and T > G transversions between primary and recurrent tumours were tested using factorial ANOVA with rank transformation. The techniques outlined in ref. 51 were followed to analyse copy number changes. Sequence quality filtering was used to remove all reads of low mapping quality (Q < 10). Due to the varying amounts of sequence reads from each sample, aligned reference reads were first used to define genomic bins of equal reference coverage to which depths of alignments of sequence from each of the tumour samples were compared. This resulted in a measurement of the relative number of aligned reads from the tumours and reference in bins of variable length along the genome, where bin width is inversely proportional to the number of mapped reference reads. A hidden Markov model (HMM) was used to classify and segment continuous regions of copy number loss, neutrality, or gain using methodology outlined previously52. The five states reported by the HMM were: loss (1), neutral (2), gain (3), amplification (4), and high-level amplification (5). In cases with germline, copy number gains and losses are called against the germline sample. In cases without germline, CNV calls were made using the primary instead of the germline sample, such that gains and losses reported in the recurrent tumour are relative to the copy number state in the primary. The limitations of this approach are that (1) when both primary and recurrent tumours share an event, the CNV output looks normal, and (2) when a gain (or loss) is called in the recurrent tumour versus the primary tumour, we cannot distinguish between the two scenarios that can give rise to such a result. The first scenario is that there is a gain the recurrence vs the primary, and the second is that there is a loss in the primary only. To resolve this uncertainty for particular chromosomes of interest in a subset of patients without germline, we additionally ran the Control-FREEC algorithm53. Control-FREEC was run using the following default parameters, with the following exceptions: breakPointType = 4, telocentromeric = 75,000, minimalCoveragePerPosition = 5. Structural variant detection was performed using ABySS (v1.3.2). Genome (WGS) libraries were assembled in single-end mode using k-mer values of k24, and k44. The contigs and reads were then reassembled at k64 in single end mode and then finally at k64 in paired end mode. Large-scale rearrangements and gene fusions were identified using BWA (v0.6.2-r126) alignments. Evidence for the alignments were provided from aligning reads back to the contigs and from aligning reads to genomic coordinates. Events were then filtered on read thresholds. Insertions and deletions were identified by gapped alignment of contigs to the human reference using BWA. Confidence in the event was calculated from the alignment of reads back to the event breakpoint in the contigs. The events were then screened against dbSNP and other variation databases to identify putative novel events. To verify SNVs, samples were subjected to targeted deep amplicon sequencing of the tumour and normal DNA. Primers were designed with the Primer3 software54 with a GC clamp and an optimal Tm of 64 °C to ensure specificity. Primers aligned against the human reference genome were tested with a combination of UCSC’s in silico PCR tool and custom in-house scripts to obtain unique hits. The primer pairs were designed such that the variant is located within a maximum of 250 bp of the 5′ or 3′ amplicon end. The primers were tagged with Illumina adapters eliminating the need for adaptor ligation during sample preparation. The Illumina adaptor tags are as follows: 5′-CGCTCTTCCGATCTCTG on the forward amplicon primer and 5′-TGCTCTTCCGATCTGAC on the reverse amplicon primers. Genomic DNA templates or library construction intermediates were used as starting material to generate PCR products using Phusion DNA polymerase (Fisher Scientific, catalogue number F-540L). The amplicons ranged in size from 188–625 bp. Amplicons were pooled by template for direct sequencing. Preparation for sequencing involved a second round of amplification (6 cycles with Phusion DNA polymerase) with PE primer 1.0-DS (5′-AATGATACGGCGACCACCGAGATCTACACTCTTTCCCTACACGACGCTCTTCCGATCTCTG-3′) and a custom PCR primer (5′-CAAGCAGAAGACGGCATACGAGATNNNNNNGTGACTGGAGTTCAGACGTGTGCTCTTCCGATCTGAC-3′) containing an unique six-nucleotide ‘index’ shown here as the letter N. PCR products of the desired size range were purified using 8% PAGE gels. DNA quality was assessed using the Agilent DNA 1000 series II assay (Agilent, Santa Clara CA, USA) and DNA quantity was measured using by Quant-iT dsDNA HS assay on a Qubit fluorometer (Life Technologies, Grand Island, NY, USA). The indexed libraries were pooled together and sequenced on the Illumina MiSeq platform with paired-end 250 bp reads using v2 reagents. An in-house generated PhiX sequencing control library was spiked in to the samples at molar ratio of 1:100. Reads were aligned using BWA-SW, and SNVs called with Samtools mpileup with the following parameters: -d 1000000 -B -C50 -DES. Indels were called using VarScan and the following parameters: mpileup2indel–min-var.-freq 0–p-value 1–strand-filter 0. SNVs with allelic frequencies greater than 15% in recurrent tumours were considered clonal. To find evidence for rare subclones (<5%) of these SNVs in the primary samples, we generated base quality (baseQ) distributions supporting the reference and all alternate alleles in the primary (and the recurrent) compartments. Due to our amplification and sequencing strategy, all reads start at the same position, and the target SNV is always at a specific position in the read (that is, a given mutation covered by 2,000 reads will be at base position 40 in all reads). Thus, unlike shotgun protocols where read starts are random, the SNVs are never affected by sequencing errors at the end of the read (where errors tend to happen more often), and cumulative sequencing error rates for whole reads are not applicable in estimating local error rates at a specific base. Instead, detection of a real mutation is only confounded by the subset of sequencing errors at the same position in the read that causes a base change to match the mutation; sequencing errors matching the other two possible bases (that is, non-reference and non-mutation) are a non-ambiguous measure of the error rate at a particular position. Thus, to distinguish sequencing errors from real subclonal mutations, for each allele (that is, the reference allele and all three alternate alleles), we generated base quality (baseQ) distributions from all reads covering the position of the mutation; the reference base was further used as the benchmark distribution of a base without appreciable sequencing errors (Extended Data Fig. 8). The non-reference alleles that had the highest (1) mean baseQ value, (2) max baseQ value, and (3) highest number of reads with baseQ values >30, were considered real events. When all three criteria were not matched, the subclonal presence of the mutation could not be confirmed. At positions where these criteria were matched, the baseQ distributions of the alternate allele closely matched the baseQ distribution of the positive control reference base, could be easily distinguished from sequencing errors, and nearly always matched the expected mutation at that position, confirming the subclonal presence of the mutation in the diagnostic sample. The allelic ratios are modelled using a binomial distribution and incorporated into the HMM Titan calculations, where the output is a list of copy number and LOH events. The Titan run for a tumour sample that has the lowest SDbw score is the optimal result and the corresponding number of clonal clusters is the optimal one—this copy number information was then chosen for use in further analysis. Minor and major copy number counts calculated from the optimal TitanCNA zygosity states were attached to the allele frequency information for each SNV and was used as input for Pyclone 0.12.3. PyClone was used to infer subclonal populations for all samples in each case. It introduces a framework that can analyse all samples from a single case in the same run improving accuracy of the inference. Pyclone outputs cellular frequencies and clonal cluster membership for each genomic position, accounting for confounding factors such as mutational genotype in the context of copy number changes. All Pyclone analyses were done using a multi-sample model and a beta-binomial distribution, with pre-calculated parental copy number inferred by TitanCNA. Copy number and LOH information was called for 14 patients with matched germline samples using Control-FreeC53, an algorithm that provides fractional copy number level for segments. Sensitive mutation calling was performed using muTect55 and clonal and subclonal somatic mutations were shortlisted if there was adequate sequence coverage in both primary and relapse tumour compartments (10 reads minimum). Shortlisted mutations and copy number segments in areas of neutral heterozygosity were used as input to EXPANDS36. Phylogenetic relationships between the subpopulations inferred by the EXPANDS algorithm in primary and recurrent tumours were generated using both SNV and copy number segments and the BIONJ algorithm. The inferred cellular prevalence values of each subpopulation was used to generate a Shannon Index value for each compartment37. We identified 14q associated genes in Shh medulloblastoma using ANOVA in the Partek Genomics Suite. Gene expression profiles were analysed according to 14q status in samples from a previously published Toronto data set containing only SHH medulloblastoma samples (n = 82)56 in a subset of cases with available SNP6 data5, 57. The top 20 ranking signature genes were applied using k-means clustering using the R2 platform (http://hgserver1.amc.nl/cgi-bin/r2/main.cgi) on a non-overlapping, independent gene expression profiling cohort from Boston58 sub-selecting only SHH medulloblastomas. Survival differences were analysed using log-rank statistics and Kaplan–Meier estimates. Two micrograms of total RNA samples were arrayed into a 96-well plate and polyadenylated (Poly(A)+) messenger RNA (mRNA) was purified using the 96-well MultiMACS mRNA isolation kit on the MultiMACS 96 separator (Miltenyi Biotec, Germany) with on-column DNaseI-treatment as per the manufacturer’s instructions. The eluted poly(A)+ mRNA was ethanol precipitated and resuspended in 10 μl of DEPC-treated water with 1:20 SuperaseIN (Life Technologies, USA). First-strand cDNA was synthesized from the purified poly(A)+ mRNA using the Superscript cDNA Synthesis kit (Life Technologies, USA) and random hexamer primers at a concentration of 5 μM along with a final concentration of 1 μg ul−1 actinomycin D, followed by Ampure XP SPRI beads on a Biomek FX robot (Beckman-Coulter, USA). The second strand cDNA was synthesized following the Superscript cDNA Synthesis protocol by replacing the dTTP with dUTP in dNTP mix, allowing the second strand to be digested using UNG (Uracil-N-Glycosylase, Life Technologies, USA) in the post-adaptor ligation reaction and thus achieving strand specificity. The cDNA was quantified in a 96-well format using PicoGreen (Life Technologies, USA) and VICTOR3V Spectrophotometer (PerkinElmer, Inc. USA). The quality was checked on a random sampling using the High Sensitivity DNA chip assay (Agilent). The cDNA was fragmented by Covaris E210 (Covaris, USA) sonication for 55 s, using a duty cycle of 20% and intensity of 5. Plate-based libraries were prepared following the BC Cancer Agency’s Michael Smith Genome Sciences Centre (BCGSC) paired-end (PE) protocol on a Biomek FX robot (Beckman-Coulter, USA). Briefly, the cDNA was purified in 96-well format using Ampure XP SPRI beads, and was subject to end-repair and phosphorylation by T4 DNA polymerase, Klenow DNA Polymerase, and T4 polynucleotide kinase respectively in a single reaction, followed by cleanup using Ampure XP SPRI beads and 3′ A-tailing by Klenow fragment (3′ to 5′ exo minus). After cleanup using Ampure XP SPRI beads, PicoGreen quantification was performed to determine the amount of Illumina PE adapters used in the next step of adaptor ligation reaction. The adaptor-ligated products were purified using Ampure XP SPRI beads, then PCR-amplified with Phusion DNA Polymerase (Thermo Fisher Scientific USA) using Illumina’s PE primer set, with cycle conditions of 98 °C 30 s followed by 10–15 cycles of 98 °C for 10 s, 65 °C for 30 s and 72 °C for 30 s, and then 72 °C for 5 min. The PCR products were purified using Ampure XP SPRI beads, and checked with a Caliper LabChip GX for DNA samples using the High Sensitivity assay (PerkinElmer, USA). PCR products with a desired size range were purified using a 96-channel size selection robot developed at the BCGSC, and the DNA quality was assessed and quantified using an Agilent DNA 1000 series II assay and Quant-iT dsDNA HS Assay Kit using Qubit fluorometer (Invitrogen), then diluted to 8 nM. The final concentration was verified by Quant-iT dsDNA HS assay. The libraries, 2×100 PE lanes, were sequenced on the Illumina HiSeq 2000/2500 platform using v3 chemistry and HiSeq Control Software version 2.0.10. Illumina paired-end RNA sequencing data was aligned to GRCh37-lite genome-plus-junctions using BWA (version 0.5.7)49, 59. This reference is a combination of GRCh37-lite assembly and exon–exon junction sequences with coordinates defined based on transcripts in Ensembl (v61), Refseq and known genes from the UCSC genome browser (both were downloaded from UCSC in November 2011; The GRCh37-lite assembly is available at http://www.bcgsc.ca/downloads/genomes/9606/hg19/1000genomes/bwa_ind/genome). BWA “aln” and “sampe” were run with default parameters, except for the inclusion of the (-s) option to disable the Smith-Waterman alignment, which is unsuitable for insert size distribution in paired-end RNA-seq data. Finally, reads failing the Illumina chastity filter are flagged with a custom script, and duplicated reads were flagged with Picard Tools (version 1.31). After the alignment, the junction-aligned reads that mapped to exon–exon junctions were repositioned to the genome as large-gapped alignments and tagged with “ZJ:Z”59. We compared the expression values (RPKM) of genes in the primary and recurrent tissues of each tumour with data in both compartments (n = 7 patients). A gene was considered differentially expression when the absolute difference between compartments was greater than 10 and the log fold-change was greater than 2. Gene sets enrichment analysis was run on differentially expressed genes that were observed in at least two patients by subgroup, using mSigDB60.
Yuen R.K.C.,Applied Genomics |
Thiruvahindrapuram B.,Applied Genomics |
Merico D.,Applied Genomics |
Walker S.,Applied Genomics |
And 31 more authors.
Nature Medicine | Year: 2015
Autism spectrum disorder (ASD) is genetically heterogeneous, with evidence for hundreds of susceptibility loci. Previous microarray and exome-sequencing studies have examined portions of the genome in simplex families (parents and one ASD-affected child) having presumed sporadic forms of the disorder. We used whole-genome sequencing (WGS) of 85 quartet families (parents and two ASD-affected siblings), consisting of 170 individuals with ASD, to generate a comprehensive data resource encompassing all classes of genetic variation (including noncoding variants) and accompanying phenotypes, in apparently familial forms of ASD. By examining de novo and rare inherited single-nucleotide and structural variations in genes previously reported to be associated with ASD or other neurodevelopmental disorders, we found that some (69.4%) of the affected siblings carried different ASD-relevant mutations. These siblings with discordant mutations tended to demonstrate more clinical variability than those who shared a risk variant. Our study emphasizes that substantial genetic heterogeneity exists in ASD, necessitating the use of WGS to delineate all genic and non-genic susceptibility variants in research and in clinical diagnostics. © 2015 Nature America, Inc. All rights reserved.