Abstracts

Abstracts for presentations are listed below, in alphabetical order, by surname.

Hayley Baird

Using Next Generation Sequencing techniques to identify Single Nucleotide Polymorphisms in Chinook Salmon

1Baird H, 1Clarke S, 1Brauning R, 1Fisher P, 2Bailey, J, 3Walker, S, 3Symonds, J
1AgResearch, Invermay Agricultural Centre, PB 50034, Mosgiel, New Zealand
2The New Zealand King Salmon Company Ltd., 10-18 Bullen St, Nelson, New Zealand
3National Institute of Water & Atmospheric Research Ltd, Bream Bay Aquaculture Park, Ruakaka, New Zealand

Understanding and improving feed conversion efficiency (FCE) in farmed New Zealand Chinook salmon is a high priority for the industry.  A research programme has been established with two main goals: 1) tank-based performance and genetic evaluation of 160 families focussing on growth, feed intake, body fat and FCE and 2) the development of a new panel of single nucleotide polymorphism (SNP) markers that will be used to search for markers linked to QTL. Since there is no salmon genome assembly available we had to come up with a strategy that did not require a reference genome. We utilised three different next-generation sequencing platforms (454, SOLiD and Illumina HiSeq) to take us straight to SNPs. After stringent filtering we arrived at 95,000 SNPs. Comparative genomics indicated that these SNPs were evenly distributed across the salmon genome. An Illumina 6K SNP chip has been developed and 4 large families genotyped for mapping and eventual QTL analysis.


Aniruddha Chatterjee

Reduced representation bisulphite sequencing indicates widespread epigenetic variation among normal individuals

Aniruddha Chatterjee1,3, Peter A. Stockwell2, Euan J. Rodger1, Ian M. Morison1,3
1. Department of Pathology, Dunedin School of Medicine, University of Otago, New Zealand  2. Department of Biochemistry, University of Otago, New Zealand. 3. National Research Centre for Growth and Development, New Zealand.

Detailed understanding of inter-individual variation in epigenetic signatures will help establish their role in altering gene expression, disease susceptibility and phenotype. We are quantifying the methylation status of almost 24,633 CpG islands (87% of all CpG islands in the genome, 18,500 of which are promoter associated) across a normal human population, by using reduced representation bisulfite sequencing (RRBS) (1) . Good sequencing results have been obtained for seven individuals (Illumina) and more libraries are being sequenced. We establised an effective bioinformatics pipeline for analysing genome-scale DNA methylation data (2). Further, new command line tool (DMAT) is being developed to detect differential methylation patterns and identify genes involved. Then by focussing on genes that show inter-individual variation, we will explore specific cohorts to document epigenetic influences on human phenotypes and disease.

1. Meissner A, et al. Genome-scale DNA methylation maps of pluripotent and differentiated cells. Nature 2008;454:766-70.
2. Chatterjee A, et al. Comparison of alignment software for genome-wide bisulphite sequence data. Nucleic Acids Research. (advance access- doi:10.1093/nar/gks150).
 Abbreviations: CGI: CpG islands, RRBS: reduced representation bisulfite sequencing , DMAT: Differential Methylation Analysis Tool


John Cleary

Calling variants in populations with pedigrees

Real Time Genomics

The data tsunami coming out of sequencing machines now enables us to compare NGS data from thousands of individuals. These populations can be from species that are important commercially and medically as well as humans (that are presumably both).   These populations are often of related individuals. A number of techniques are available for leveraging this population and pedigree information to obtain either high quality variants or good quality ones using very low coverage NGS data.  Of particular note here is the need to provide good statistical confidence when attempting to extract differential calls such as de novo mutations. Bayesian algorithms for such applications will be described along with results on humans and other species. These include:  improvements in the quality of variant calls given population pedigrees and how this varies with the coverage; detection of de novo mutations;  detection of potentially causative alleles for traits; and detection of mutation in tumors when compared with normal cells. In the most extreme cases I will show it is possible to get a good set of variant calls for an individual where no sequencing has been done at all.


Lesley Collins

NGS project analysis on small-scale computing (aka You did WHAT on a laptop?)

Institute of Fundamental Sciences, Massey University, New Zealand

Benchtop NGS sequencing is now offering researchers a chance to answer genome-wide questions on smaller projects (and smaller budgets). However, these projects often do not have access to large computer servers or expert bioinformaticians without spending large parts of their research budget, so researchers try to do it themselves.  The most common question that these DIYers ask is: How big does my computer have to be to get things done? To address this question some publically available data was downloaded and run through two typical analysis pipelines on available desktop and laptop computers.  This talk will present results to show what is at present possible for small-scale NGS analysis and where common problems lie. Of course with all bioinformatics, if symptoms persist, please seek professional advice.

Christine Couldrey

Development of epigenomic pipelines for use in agricultural animals
 
Christine Couldrey, Rudiger Brauning, Ernest Retzel, John McEwan
AgResearch
 
The creation of single nucleotide polymorphism chips and the development of genome wide selection will allow the animal breeding industry to take a quantum leap in the rate of genetic progress. However, soon all sequence variations in each individual in closed breeding schemes will be known. What is not currently known is how to rank these variations, especially those that involve changes in gene expression rather than amino acid sequence. One of the key determinants in the control of gene expression in mammals is DNA methylation – a mechanism known to play a central role in regulating many aspects of growth and development. High-throughput sequencing has recently become a vital tool in the analysis of DNA methylation and reduced representation bisulfite sequencing (RRBS) has proven to be effective in understanding DNA methylation landscapes. However, to date, mammalian genome wide epigenetic studies have focused on humans and mice. Here we describe development of RRBS in sheep. The Carwell phenotype, used as a proof of principle, is a desirable inherited muscular hypertrophy for which, in spite of considerable resequencing efforts, the causative mutation has not been identified. Muscle DNA was used for RRBS and epigenomic pipeline development to generate single nucleotide resolution analysis of DNA methylation in sheep.  Sequence read quality was assessed and displayed the expected nucleotide composition. Analysis of 1% of the genome resulted in analysis of >20% of all CpG sites.  DNA methylation analysis was precise with biological replicates showing high repeatability (r>0.9). Methylation measurement was accurate as illustrated by the comparison of RRBS methylation data and the gold standard Sequenom analysis within the Carwell region where proportion methylation measured matched at 132/134 CpG. Sheep were different from other species; as silico analysis of the sheep genome highlighted greater CpG island enrichment by RRBS than expected. We have generated the first sheep methylome and optimised RRBS for use in agricultural animals.  This protocol and bioinformatic pipeline will have widespread use in the future – by facilitating the identification of superior individuals to enhance productivity through further selective breeding.


Brian Fritz

Current and imminent state of the art for Illumina’s sequencing technologies

Illumina

Illumina develops, markets and supports innovative and integrated sequencing technologies, applications and data analysis solutions for genetic and genomic analysis. Our tools and services accelerate genetic analysis research and fuel advances in consumer genomics, diagnostics, plant and animal genetics and many additional applied genetics markets. In this presentation we will outline the current and imminent state of the art for Illumina’s sequencing technologies including the current output specifications for Illumina’s sequencing instruments.  We will then detail Illumina’s current end-to-end workflows, from sample preparation through sequencing to analysis, for diverse DNA, RNA and epigenetic research applications.


Chad Harland

Pipelines, Pedigree & Polymorphism

Chad Harland, MSc (BCHM), Science Intern, Research & Development, LIC, Hamilton

A variety of different bioinformatics pipelines exist for NGS datasets, each with slightly different trade offs. Having tested a number of these pipelines for the DairyPGP project we have settled on a mixed solution that offers high performance, high sensitivity and specificity for SNP and Indel detection.

Secondly due to the population structure of the NZ Dairy Herd we have a significant amount of Pedigree information that can be used to improve Variant calling. As such we have looked at the use of trios, pedigree and Mendelian inheritance of variants in Variant calling and filtering and will discuss the tools used and the advantage this additional information provides.


Nick Heng

De novo sequencing of the genome of Streptococcus trichosurus, a new oral bacterial species isolated from the New Zealand brushtail possum Trichosurus vulpecula

Nicholas Heng1*, Ceridwen Benn1, John Hale2, Jules Kieser1, Jo-Ann Stanton3

(1) Sir John Walsh Research Institute (Faculty of Dentistry), University of Otago, Dunedin, New Zealand, (2) BLIS Technologies, Dunedin, New Zealand, (3) Department of Anatomy, University of Otago, Dunedin, New Zealand.

Members of the bacterial genus Streptococcus inhabit a multitude of sites in humans and many animals. Whilst some species are pathogenic, most are commensals. During a recent survey of oral streptococci from New Zealand brushtail possums (Trichosurus vulpecula), a new species (provisionally called Streptococcus trichosurus) was identified. The genome of S. trichosurus was sequenced using the new Ion 318™ sequencing chip in combination with the Life Technologies Ion Torrent-based Personal Genome Machine, yielding ~485 Mbp (~210-fold coverage) of sequence data. Here, the draft S. trichosurus genome sequence is presented and the bioinformatic procedures associated with its assembly are discussed.


Mike Keehan

An update on year two of the PGP Dairy Genomics Sequencing project

LIC

The second year of the Dairy PGP project will have seen LIC and Vialactia receive an additional 11 TBase of whole genome sequence data. Preliminary results from applying this tranche of data to phase and impute whole genome sequence from 50K SNP chips will be presented. Practical experiences from year one and two will be summarised.  The state of the bovine genome assembly will be discussed. The project offers an opportunity for biologists who would like to examine a large whole genome population dataset.   


Lia Liefting

Diagnosis of Plant Pathogens using Next-Generation Sequencing

Lia Liefting, Wellcome Ho and Lisa Ward
Plant Health and Environment Laboratory, Ministry for Primary Industries, P.O. Box 2095, Auckland 1140

MPI’s Plant Health and Environment Laboratory is responsible for the identification of new pests and diseases affecting plants, plant products and the environment.  In some cases identification can take an extended period, especially for new pathogens and emerging diseases.  Metagenomic analysis using next-generation sequencing (NGS) has the potential to detect the full spectrum of pathogenic organisms in a single test.  Therefore this technology is being implemented in our laboratory for plant pathogen diagnosis.  Our strategy is to use total RNA in order to detect all pathogen types.  The processes involved in the preparation of the RNA for NGS will be described as well as validation of the method on at least one representative from each of the major groups of plant pathogens and genome types.

 

Prof Si Lok

A Small Genome Centre’s Adoption of New Generation DNA Sequencing for Research and CORE Service

Si Lok, PhD, Professor of Practice in Applied Genomics, Li Ka Shing Institute of Health Sciences, The Chinese University of Hong Kong, SAR China.

New generation DNA sequencing offers unique opportunities and challenges to a small genome centre.  At our centre, we focus the use of the technologies and expertise for difficult collaborative projects not generally suited for the routine commercial service providers.  The present talk highlights the latest technologies, methodologies and applications in the field as well as some of our centre’s efforts.  We compare solution-based hybridization enrichment and high throughput Amplicon-based targeted resequencing of exomes or candidate genes. The newly established RainDance platform for massively parallel amplicon generation combined with 454-pyrosequencing is particularly powerful for small targeting studies. Other research efforts are in the areas methylomics, transcriptomics, and pathogenic genomics where the power of DNA sequencing is used to identify emerging pathogens and to study host and pathogen interactions.

 

Bennet McComish

Index-free de novo assembly and deconvolution of mixed mitochondrial genomes

Bennet J McComish, Simon F Hills, Patrick J Biggs and David Penny

In order to make use of the high throughput available with next-generation sequencing technology, we developed a pipeline for sequencing and de novo assembly of multiple mitochondrial genomes without the costs of indexing. We first used simulations to explore
the ability of existing sequence assembly algorithms to separate and assemble sequences from different sources. Once optimised, the same methods were successfully applied to reads from a single lane of an Illumina Genome Analyzer flow cell containing a mixture of PCR products from six different mitochondrial genomes. More recently, we applied a modified version of the same pipeline to four more
mixtures, this time using total genomic DNA, and successfully assembled 17 mitochondrial genomes.


Alan McCulloch

A method for designing NGS transcriptomics experiments that  includes in-silico biological replication

AgResearch

An NGS transcriptomics experiment usually involves alignment of the sequenced transcripts to a reference ‘ome (genome or transcriptome), in order to identify and quantify the expressed loci. However a given reference ‘ome is but a single sample from a large space of potential alternative assembled ‘omes. Because we only take a single sample from this ‘ome space (either we choose an ‘ome assembled by somebody else, or we assemble one on the fly ourselves), the experiment provides no information at all about the influence our sample of reference ‘ome has on NGS expression observations, yet there is evidence that reference ‘ome assemblies can vary significantly due to both technical factors and underlying genetic variation.  While it is straightforward to align transcripts to multiple references, establishing the homology relationships between the references needed to compare the results is difficult, and this cannot be fully automated, so that this type of in-silico biological replication is seldom if ever included in experimental designs. Here we suggest a method for designing NGS transcriptomics experiments that includes in-silico biological replicates, in a way that could be automated and included as part of a standard NGS transcriptomics pipeline. This design would deliver stabler NGS transcriptomics assay results robust against common variations in reference `ome, and may extend the applicability of NGS transcriptomics to more highly variable species, in which NGS expression observations obtained from alignment to a single reference `ome may be too unreliable to be useable.


Martin Kennedy

Earthquake induced stress cardiomyopathy: is it a Mendelian condition?
Cameron Lacey1, Paul Bridgman2, Vicky Cameron3, Roger Mulder1, Julie Zarifeh4, Mik Black5, Murray Cadzow5, Tony Merriman5, Alan Aitchison6, Martin Kennedy6.

1Department of Psychological Medicine University of Otago, Christchurch; 2Cardiology Dept., Canterbury District Health Board, Christchurch; 3Department of Medicine, University of Otago, Christchurch; 4Psychiatric Consultation Service, Canterbury District Health Board, Christchurch; 5Biochemistry Department, University of Otago, Dunedin; 6Department of Pathology, University of Otago, Christchurch

The major earthquakes of 4th September 2010 and 22nd February 2011 both triggered case clusters of a rare condition called stress cardiomyopathy (also known as broken heart syndrome or Takotsubo cardiomyopathy). Many of these patients received critical care in the coronary care unit of Christchurch Hospital, and some required intensive care with ventilatory support, but ultimately all survived. The resulting very well characterised, tightly homogenous cohort of 30 patients is unprecedented. Almost all patients presenting with the condition were post-menopausal females, consistent with other reports (Wittstein et al.) . This provides a unique opportunity to study the underlying causes and presentation of this perplexing disorder.
The exact aetiology of stress cardiomyopathy remains unknown with catecholamine induced myocardial stunning a proposed pathway (Summers et al.). Many forms of cardiomyopathy have genetic origins, and it is reasonable to propose that this syndrome arises from a very rare underlying genetic predisposition that is exposed in times of major, acute stress. We hypothesised that stress cardiomyopathy is a rare Mendelian predisposition that is exposed with acute major stress. The rarity of the underlying mutation requires that large numbers of people must be exposed to the stressor, which only happens in times of natural disaster such as major earthquakes. This is rather speculative, although two prior reports describe occurrence of the syndrome in relatives (Kumar et al., 2010; Pison et al., 2004) . Exome sequencing provides a method to test this hypothesis. 
We obtained exome data on 12 of the Christchurch earthquake stress cardiomyopathy patients using Illumina TruSeq exome enrichment and the Illumina HiSeq platform (New Zealand Genomics Ltd). The data have been processed through the GATK pipeline, and we are examining candidate variants that occur in patient samples with a higher than expected distribution based on 1000 Genomes Project data. This presentation will describe these preliminary analyses and some of the pitfalls encountered so far.

Kumar G, Holmes DR, Jr., Prasad A (2010). "Familial" apical ballooning syndrome (Takotsubo cardiomyopathy). Int J Cardiol 144(3): 444-44

Pison L, De Vusser P, Mullens W (2004). Apical ballooning in relatives. Heart 90(12): e67.

Summers MR, Lennon RJ, Prasad A (2010). Pre-morbid psychiatric and cardiovascular diseases in apical ballooning syndrome (tako-tsubo/stress-induced cardiomyopathy): potential pre-disposing factors? J Am Coll Cardiol 55(7): 700-701.

Wittstein IS, Thiemann DR, Lima JA, Baughman KL, Schulman SP, Gerstenblith G, et al. (2005). Neurohumoral features of myocardial stunning due to sudden emotional stress. The New England journal of medicine 352(6): 539-548.


Mary Morgan-Richards

Hybrid origin of a parthenogenetic genus: the genomic evidence

Mary Morgan-Richards1, Simon F. K. Hills1, Patrick J Biggs2 & Steve A Trewick1

  1. Institute of Natural Resources Massey University, Palmerston North 4442, New Zealand
  2. Institute of Veterinary, Animal & Biomedical Sciences, Massey University, Palmerston North 4442, New Zealand

Hybridization between species can combine divergent genomes and create new species when reproductive isolation from parentals accompanies the novel genome fusion (Bullini 1994). It has been estimated that approximately 70% of plants are the result of allopolyploidy. Hybrid species can be recognized by the presence of alleles distinct to two species co-occurring in the same genome.

A hybrid origin for an endemic New Zealand genus of stick insects (Acanthoxyla) was suggested (Morgan-Richards & Trewick 2005) with the related bisexual species, Clitarchus hookeri, named as a putative paternal species.  A maternal bisexual species has not been identified and is likely to be extinct (Trewick et al. 2008; Buckley et al. 2010).  It is also likely that some lineages of Acanthoxyla are triploid, and it is possible that Clitarchus hookeri was not involved in the origin of all Acanthoxyla species (Buckley et al. 2008; Myers et al. unpub).

Next generation DNA sequencing provides large datasets for testing hybrid origin hypotheses and here we set out a procedure for evaluating such data. Using de novo assembled transcripts to compare ‘allelic’ diversity in putative hybrids and their putative parents, we have used mRNA sequences to examine the allelic diversity within one Acanthoxyla lineage and compared this to homologous gene sequences from Clitarchus hookeri. The hybrid origin hypothesis predicts that at each locus Acanthoxyla will contain an allele similar to that of Clitarchus hookeri, and one allele from the unidentified maternal ancestor.  If Acanthoxyla is not of hybrid origin then the two alleles within Acanthoxyla will be more similar to each other than either is to the Clitarchus hookeri alleles.  We also present evidence to address the questions: Is Acanthoxyla diploid or triploid? And does Acanthoxyla use apomictic or automictic parthenogenesis to reproduce?


Kristina Ramstad

HiSeq transcriptomics to discover genes associated with reproductive success in kiwi

KM Ramstad and G Kolle

Little spotted kiwi (LSK)and rowi have the smallest population sizes and lowest neutral genetic diversity of the five currently recognized species of kiwi (Family Apterygidae).  Both LSK and rowi exhibit low hatching success and high variance in reproductive success. The latter effect is particularly dramatic in rowi where a full third of adult birds do not breed.   Poor reproductive performance may be due to historical bottleneck effects and the subsequent inbreeding effect of small population size. We have performed RNA-sequencing of 16 individual kiwi across the two species from whole blood isolates.  As there is no current genome, we performed transcriptome de novo assembly using a subset of the read data.  The de novo assembly output identified 13,000 unique protein coding transcripts with homology to human and/or chicken.  More than 3,000 of these transcripts are predicted to cover the full length of the ORF and a majority of the remainder nearly full length.   Realignment of each species against the reference easily identified numerous SNPs/indels, with high confidence that they represent species and individual specific markers.  These mutations are being evaluated for changes to known breeding genes and will act as markers for determining genetic variation among individuals with varying reproductive success.


Richard Spence

New Zealand Ostreid herpesvirus – application of high throughput sequencing in a biosecurity context

Suzi Keeling, Richard Spence, Cara Brosnahan, Colin Johnston, and Wendy McDonald

Bacteriology and Aquatic Animal Diseases Team., Investigation and Diagnostic Centre, Compliance and Response, Ministry for Primary industries

In November 2010 the Ministry for Primary Industries (formerly MAF) was notified of high mortality levels in juvenile Pacific oysters in the North Island of New Zealand.  Ostreid herpesvirus was identified in association with the mortalities.  Ostreid herpesvirus has plagued the European Pacific Oyster industry for over a decade resulting in significant economic losses.  Although it appears the virus has been present in New Zealand for several years this was the first significant mortality event associated with the virus in New Zealand.  A metagenomic analysis of highly infected oyster larvae was undertaken using the Roche GS Junior sequencer to try and facilitate a better understanding of factors contributing to the mortality event.  Primary analysis of the data resulted in assembly of an Ostreid herpesvirus genome against a reference genome.  More detailed analysis revealed that the New Zealand Ostreid herpesvirus harboured several significant deletions in comparison to the reference genome.  In addition, de novo assemblies performed on the data identified the presence of a number of Vibrio species that may also have contributed to the observed mortalities.  Further analysis of this data set is focused on the observed differences between the New Zealand Ostreid herpesvirus and the reference genome and what impact this may have on the pathogenicity of the strain.  


Peter Stockwell

Differential Methylation Analysis using RRBS: Challenges and New Insights

Peter A. Stockwell2, Aniruddha Chatterjee1,3, Euan J. Rodger1, Ian M. Morison1,3
1. Department of Pathology, Dunedin School of Medicine, University of Otago, New Zealand  2. Department of Biochemistry, University of Otago, New Zealand. 3. National Research Centre for Growth and Development, New Zealand.

Reduced Representation Bisulphite Sequencing (RRBS) resolves DNA methylation status at base-pair resolution and enriches for promoter associated CpG islands (CGI) in the genome.  Since CGI methylation plays a key role in the epigenetic regulation of gene expression, there is interest in elucidating differential methylation patterns of CGIs and neighbouring genes in normal individuals and in disease conditions.  We describe the unique challenges associated with DNA methylation data analysis and our efforts in quantifying differentially methylated regions in RRBS contexts and in relating them to known genomic elements.  The challenges relate to the nature and volume of data, the sizes of genomes and the complexity of scanning for differential methylation of RRBS fragments which are discontinuously distributed along the genome.


Tracey van Stijn

High-throughput Genotyping-by Sequencing (GBS) in Sheep

Tracey van Stijn, Shannon Clarke, Rudiger Brauning and John McEwan, AgResearch

Recent advances in next generation sequencing technology have increased the output/cost to a level that now allows for the potential of GBS in livestock.  We have explored GBS in sheep with the aim of developing a cost effective, reproducible and high-throughput SNP genotyping method that can be manipulated to return varying genome coverage.  Restriction enzymes have been employed together with adapter based multiplexing for next generation sequencing, a technique that has been well established for high-density SNP discovery and genotyping in numerous species.  We utilised the GBS method described by Elshire et al., 2011 (PLoS ONE 6:e19379), however, through the addition of specific nucleotides within and following the restriction enzyme cut site to the PCR primer we are able to reduce the complexity of the genome in a controlled manner.  Varying the number of samples and the ‘size’ of the reduced genome per lane on an Illumina HiSeq2000 allows for differing magnitudes of SNPs to be genotyped and interrogated.  The method for reducing the complexity of the genome, sequencing and bioinformatics pipeline for GBS in sheep will be presented.


Steve Wylie

Tailoring high-throughput sequencing approaches to next-generation plant virology in south-west Australia.

Steve Wylie1, Hua Li1, Jamie Ong1, Kingsley Dixon2, Aaron Tan1, Mike Jones1

1 Plant Virology Group, Western Australian State Agricultural Biotechnology Centre, Murdoch University, WA 6150, Australia
2 Botanic Gardens and Parks Authority, Kings Park and Botanic Garden, West Perth, Western Australia 6005, Australia.

Given the great age of the Australian continent, high endemism amongst its plant flora, and the varied ecosystems present, we hypothesised that its indigenous viral flora would be much richer than that currently described.  Many exotic plants and virus vectors have become established in Australia over the past two centuries, and we also predicted that the indigenous flora would be suffering invasion by aggressive exotic viruses inadvertently imported in these new species.  We are testing these hypotheses on a range of indigenous and exotic plant groups, with a focus on terrestrial orchids, a group of conservation concern.  
Approaches used to identify RNA viruses include sequencing RNA from single plants; sequencing RNA pooled from multiple plants coupled with subsequent matching of plants and viruses found; sequencing small RNA species generated by plants in defense of viruses; labeling RNA from individual plants before pooling and sequencing; and developing methods to enrich plant RNA for viral transcripts before sequencing.  The pros and cons of the various approaches tested will be discussed, as will implications of this work on conservation management and biosecurity policy.