- Our Vision
- UGI Newsletter
- MSc Courses
- UGI Core Services
- Gene Ontology
- FH Variant Databases
- UGI Calendar
- Contact Us
- Bloomsbury Centre for Genetic Epidemiology and Statistics
Seminar 26 Sept: Modelling rate of decline (slope) of a biomarker as a quantitative genetic trait using mixed effects models: application to loss of renal function (GFR) in type 1 diabetes
UGI Newsletter July 2013
Welcome to the UGI Spring 2013 newsletter. Please contact Susmita Datta if you have any comments or information to be included in the next edition.
- New members of UGI
- Bloomsbury Centre for Genetic Epidemiology and Statistics
- UGI MSc Courses
- Other news & information
- UGI distribution list & Facebook page
- Recent publications
Congratulations to Doug Speed for being named the best FMT-BIR Young Biometrician of the year!
Photo L-R: Professor David Spiegelhalter, Professor Stephen Senn, Doug Speed and Professor Simon Thompson
Doug has been named the best young biometrician in the UK by the Fisher Memorial Trust and Biometric Society. Stephen Senn, the Secretary of the Fisher Memorial Trust, presented the award to Doug Speed on 3rd July 2013 before David Spiegelhalter gave the keynote talk.
"The unanimous decision of the judges is that the prize should be awarded to Doug Speed (UCL) for "Improved heritability estimation from genome-wide SNPs" in Am J Hum Genet (2012). This elegant paper uses critical thinking and seemingly simple mathematics to propose a simple adjustment for linkage disequilibrium when testing for association between SNPs and diseases. Together with related software, it has the potential to make a large impact in post-genomic genetics."
Welcome to new members of UGI
Oliver joined UGI in January this year to establish a new lab that collects, analyses and visualizes data from large population samples around the world to explore the dynamic relationship between genotype and phenotype in humans.
Genetic and environmental variation affects all complex human traits and disorders. However, which variants affect us and how they have their effects depends on environmental and developmental context. For example, some genetic effects may be apparent only in childhood, or only in the centre of large cities. The group uses approaches including twin model-fitting and genome-wide analysis in samples such as the Twins Early Development Study (TEDS) of 10,000 pairs of UK twins (http://www.teds.ac.uk), as well as developing new statistical and visual analysis approaches to big data. Recent projects from the lab include studies showing increasing genetic influence across time for cognitive abilities, and the spACE project which is mapping the UK’s genetic and environmental hotspots: http://bit.ly/tedsgeo. Previously Oliver read Natural Sciences at Cambridge, specialising in Experimental Psychology. He went on to study for an MSc and PhD in Statistical Genetics at the MRC Social, Genetic & Developmental Psychiatry Centre, King’s College London. In the final year of his PhD he was awarded a Sir Henry Wellcome Fellowship by the Wellcome Trust to spend the next four years developing ways to explore patterns of genome-wide association in the development of cognitive, behavioural and psychiatric disorders. During his fellowship he spent time at the Wellcome Trust Centre for Human Genetics in Oxford, and at the European Bioinformatics Institute in Cambridge.
Manuela has recently joined the UGI as a Postdoc in Vincent Plagnol’s group. Her current work focuses on understanding the molecular mechanisms underlying ocular inflammatory and autoimmune diseases. Manuela holds an MEng in Electronic Engineering from the University of Cagliari (2005) and a PhD in Computer Science from the University of Manchester (2010), from where she gained expertise in machine learning and information theory. Following her PhD she took a postdoctoral position in Matt Hurles' group at the Wellcome Trust Sanger Institute, where she worked on a genome-wide association study investigating the role of copy number variants in Type I diabetes. Her main interests involve the development of computational and statistical methods for the analysis and the integration of next-generation sequencing data in order to gain insights into disease genetics.
We would like to say goodbye to the following members of staff:
Margarida Lopes who will soon be leaving to start a new job with Illumina, as well as to Sonia Shah who recently left to start a post-doc in Peter Vischer's group in Australia.
Bloomsbury Centre for Genetics Epidemiology and Statistics (BCGES)
The Bloomsbury Centre for Genetic Epidemiology and Statistics held its
Annual Scientific Meeting on 11th June. This year, the meeting was
organised together with the South of England Genetic Epidemiology Group (SEGEG)
and speakers included Ian Dunham (WT Sanger Institute) and Vardhman
Rakyan (Queen Mary UL).
The Centre will run two short courses in September 2013.
Course 1: Introduction to Genetic Epidemiology in the
(3rd – 6th September 2013)
Genetic epidemiology holds great potential for personalised medicine and improved biological knowledge of disease processes. This course provides an introduction to the design, analysis and interpretation of genetic studies of disease, with a focus on state of the art analysis of genomewide association scans. Throughout the course participants will gain practical experience of analysing genetic data in population and family studies. By the end of the course participants will have an understanding of the fundamental concepts of genetic epidemiology, will have a working knowledge of the terminology and current status of the field, and will be able to perform many basic analyses of genetic data.
For more detailed information and to apply please visit - course booking.
Course 2: High throughput sequencing in disease studies
(9th – 12th September 2013)
Rapidly developing technologies now allow genomes to be sequenced more quickly and cheaply than ever before. This course will cover state of the art methods and applications of next generation sequencing. The course runs over 4 days and participants will be introduced to tools for analysing high throughput sequence data, including methods for measuring copy number variants and allele-specific expression, and conducting disease association analysis with sequence data. There will be considerable opportunities to gain practical experience with new data types such as whole genome sequence, RNA- and ChIP-seq data. By the end of the course participants will have a broad knowledge of the state of the art and will be well equipped to analyse their own data.
For more detailed information and to apply please visit - course booking.
If you would like to join the e-mail list that will announce future seminars in statistical genetics and genetic epidemiology and other related events of broad interest, you can subscribe here.
UGI MSc courses
Please contact Elvira Mambetisaeva at firstname.lastname@example.org if you have any comments or information regarding the UGI MSc courses to be included in the next edition.
In May, all our students from both MSc courses on the Genetics of Human Disease and on Pharmacogenetics and Stratified Medicine courses sat their module exams. They will be working on their MSc projects until mid-August. As always we were able to offer our students a wide range of very exciting laboratory and computer-based projects supervised by researchers from a number of UCL Institutes, including the Institute of Child Health, the Cancer Institute and the School of Pharmacy. Students’ project presentations will take place on 17th-18th September 2013. All welcome!
New MSc programme in Computational and Genomic Medicine
UGI is in the process of setting up a new MSc programme on Computational and Genomic Medicine starting in September 2014. The MSc steering committee designed a stimulating programme content which has the strong support of UCL staff and undergraduate students which has been already approved by the Department of Genetics, Evolution and Environment and the Division of Biosciences.
The new MSc aims not only to provide students with general knowledge of Bioinformatics but also to equip them with the specialised knowledge and skills required to use post-genomic data for predicting and defining the genetic basis of various human diseases. As well asto allow students to be able to analyse human genomic data to develop innovative diagnostic tools and therapeutic approaches. It will provide an integrated view of computational and genomic science research, with modules in Biocomputing (held at Birkbeck College), Gene Ontology, Statistical Computing, Genomic Analysis, Bioinformatics and Translational medicine, as well as providing technical, in-depth knowledge and skills of different research techniques in these fields. Through choice of optional modules, students will be able to focus on Molecular Evolution, Computational and Systems biology and Statistics for interpreting genetic data. We believe that this will be a popular and important addition to the Masters’ portfolio of the UCL Faculty of Life Sciences.
News from our former students
We are always delighted to hear from our former students and learn about their achievements. Marilena Mela, a former student from the MSc in Genetics of Human Disease, sent us her greetings from the University of Southern California where she was offered a job as a research laboratory specialist with further opportunities to conduct a PhD there.
Mark Bartlett who completed the MSc in Pharmacogenetics and Stratified Medicine programme last year has set up Geneix, a London-based health-tech startup. Mark and the cofounder Adam Harman-Clarke have developed InterAct, an app to screen new medications against a patient’s current prescriptions and, for the first time in the UK, against a patient’s genetic profile as well. You can find more about Geneix and InterAct by visiting www.geneix.co.uk.
Mark’s other big news is that the results of his MSc project were published recently in the Lancet (Vol.381; Number 988, Jun 01, 2013). The paper discusses the results of the a quantitative online survey which was design to assess whether physicians' opinions are acting as a barrier to the adoption of genetic testing in drug prescription processes in the UK. Mark was supervised by Professor Elizabeth Shephard from the UCL Institute of Structural and Molecular Biology, Division of Biosciences. You can access the article via the following link:
to the Masters courses for 2013-14
If you are still looking for an exciting postgraduate programme in Biomedical Sciences then why not to look for our interdisciplinary MSc in Genetics of Human Disease and the MSc in Pharmacogenetics and Stratified Medicine programmes at our website http://www.ucl.ac.uk/ugi/education. We welcome our UK students to apply and the deadline for applications is 3rd August 2013. There are 4 fully funded studentships available to UK students so hurry up and apply! Yuo cannot be considered for the scholarship without first being accepted on the course! We believe that you will find the Division of Biosciences at UCL to be a stimulating and invigorating environment in which to carry out your postgraduate study. We also believe that that these programmes will improve your career prospects.
Further details regarding our new MSc Computational and Genomic Medicine
will be up very soon for a first intake in 2014, so watch this space!
Other news & information
Report from 4th International Biometrics Society Channel Conference 2013, St Andrews (by Doug, Dace, Orlando and David - pictured below)
Report from 4th International Biometrics Society Channel Conference 2013, St Andrews (by Doug, Orlando, David & Dace - pictured below)
The biometricians of North West Europe descended on Scotland this July, and were joined by four members of the UGI. Lasting three days, there was a wide span of talks, ranging from Systems Ecology to Medical Statistics; the talks reflected the breadth of quantitative genetics nowadays, but also how in recent years the focus has move from wildlife and ecology to humans. The keynote speaker was David Spiegelhalter, who presented on the pitfalls of communicating statistics to the public. In particular, he discussed the problems in making relative risks understandable and interpretable to the person on the street. For example, what does it mean when a statistician tells us that eating five fruit and veg a day reduces an individual's "instantaneous death hazard" by 20%, while having a bacon buttie every day increases it by 10% - how should those figures be made interpretable? The proposed solution is to discuss choices in terms of their effect on an individual's effective age - so a 30 year old who follows the five-a-day advice would have the effective age of a 28 year old, but were the same person to instead sneak to the greasy spoon each morning, he would have the effective age of a 31 year old - when framed as a three year difference in life expectancy, it should become easier for Average Jo to make an informed decision.
Spiegelhalter's talk, however, was exceptional for lacking any mathematical symbols, as the majority of presentations were filled with pages of complex formulae. A hot topic remains searching for effective multivariate regression methods. There was continued dispute between the sparsity and shrinkage advocates - the former, who favour methods such as the lasso, seek models where only a few predictors contribute, while the latter believe it's best to include all predictors in the model, but shrink each predictor's effect size to zero. There remains no outright winner. From within the UGI, Dace Ruklisa presented her work on efficiently distinguishing between pathogenic and harmless human sequence variations. Dace showed that by incorporating conservation information and characteristics of the genes local to the variants, her Bayesian logistic method was able provide both better sensitivity and specificity compared to the classifiers currently being used. Orlando Doehring outlined the need for tools to accurately determine the breed of dogs - knowing that a dog is a pure or designer breed can increase its value by hundreds of pounds, but also for mixed-breed dogs, authorities are keen for ways to identify if a dog is made up of any breeds considered dangerous. Orlando presented his work, which uses state-of-the-art Markov Chain Monte Carlo inference to estimate a mixed-breed dog's composition. He demonstrated that his method was better able to determine a dog's component breeds, and could even be effective when all eight of a dog's great-grandparents were of different breed.
The session on "Omics" emphasised that while much attention is placed on understanding how sequence variations affect phenotypes, there are many other (perhaps less obvious) factors which might have an influence. Two examples are glycans, which measure sugar levels within cells, and microbe composition, as each human cell will contain a mosaic of genetic material from bacteria (as well as material from many other foreign species). However, the many difficulties in accurately measuring and quantifying these alternative factors, makes it much harder to determine their phenotypic impacts. Finally, those who think analysing human data is difficult, should spare a thought for those working on other species with inherently more complex structures. For example, sugar cane is not limited to a "ploidy of two" (as is the case in humans, where chromosomes almost always exist in homologous pairs). Instead, the ploidy of sugar cane can be as high as eight, and will vary depending on which chromosome is being considered. So while many talks appreciated the benefit of exchanging ideas and methods across fields, a number of problems, such as genotyping sugar cane, still call for species-specific solutions.
Sense about Science - update from David Balding
Access "Sense About Genetic Ancestry Testing" guide which has had a big input from David and Mark Thomas. The guide tries to debunk some of the nonsense that the public is often told about genetic ancestry testing. There's been quite a bit of press and TV coverage, here's the UCL and BBC websites:
and two recent press stories:
Your tongue can tell of a Viking past
Are you related to Cleopatra? Or are genealogists fishing in the Nile?
UGI along with the Faculty of Life Sciences hosted a conference on Next Generation Sequencing which was held in March at the Institute of Child Health. The event was very well attended and main speakers included Ewan Birney from the EBI and Jeffrey Barrett from WTSI.
In May this year, we also held another Beer & Pizza Science evening, this time the topic was Genetics in Drug Metabolism and Pharmacogenetics.
UGI email distribution list
If you would like to subscribe to receive information
about events organised by UGI please visit UGI email list.
UGI is on Facebook
The Institute has a Facebook page where you can find all the latest information and events taking place. Visit us at www.facebook.com/uclgeneticsinstitute
Recent papers and publications
Seven new loci associated with age-related macular degeneration - UGI's Valentina Cipriani publishes in Nature Genetics
Prof Philippa J Talmud DSc , Sonia Shah MSc †, Ros Whittall MSc, Marta Futema BSc , Philip Howard BSc , Jackie A Cooper MSc, Seamus C Harrison MRCS, KaWah Li MSc, Fotios Drenos PhD, Frederik Karpe PhD, Prof H Andrew W Neil DSc, Olivier S Descamps PhD, Claudia Langenberg PhD, Nicholas Lench PhD, Prof Mika Kivimaki PhD, Prof John Whittaker PhD, Prof Aroon D Hingorani PhD, Meena Kumari PhD, Prof Steve E Humphries PhD
Evaluation of mixed-source, low-template DNA profiles in forensic science - David Balding publishes in the online eddition of the Proceedings of the National Academy of Sciences of USA. The article describes David's software for low template forensic DNA profiles, and its application to some simulated and real crime scene data, including from the case of Knox and Sollecito in Italy.
Marco Scutari has recently published a book titled Bayesian Networks in R, futher information on the book can be found here.
Page last modified on 12 jul 13 14:25