PhD Student 'targeted and personalized cancer therapy development by high-dimensional data analysis' (36hours/week)
Department of Epidemiology and Biostatistics, VUmc, Amsterdam
Use your quantative modeling skills to enable development of new statistical models and analysis tools and help improve targeted and personalized therapy.
The project
Joint statistical analysis of multiple -omics data can unravel genomic networks but, as it involves more variables than samples, requires new methods. Previously we developed a powerful model to test for associations (the 'needles') between genome-wide gene expression and copy number (the 'haystack'), producing lists of genes regulated by copy number change (Menezes et al., BMC Bioinformatics, 2009). However, methods to find types of regulatory mechanisms still lack, such as when two or more molecular mechanisms work in different ways to affect gene expression. Understanding regulatory mechanisms in cancer is especially important for targeted and personalized therapy development. For more information about our department, see http://tinyurl.com/EpidemiologyAndBiostatistics
Your challenge
As a PhD Student your main tasks and responsibilities are:
• Develop statistical methods and tests to find specific types of gene expression regulation, using multiple high-dimensional datasets simultaneously;
• Develop methods, based on tests and/or estimation, that can find specific types of joint effects of covariate sets (molecular profiles) on the response (gene expression);
• Analyse the data and write scientific publications which lead to your thesis.
Your profile
We are searching for a PhD Student who has the following expertise:
• A MSc in either Mathematics, Statistics, Bioinformatics or a closely related discipline;
• You have a strong background in quantitative modeling including, ideally, either statistics or mathematics;
• You have programming skills, preferably in R;
• You can explain complex ideas to biologists/non-specialists;
• You have good communication skills, you are self-disciplined and you can work in a multi-disciplinary environment.
Benefits
Salary Scale: OIO (EUR 2.200 in the first year that increases to EUR 2.818 gross per month in the fourth year, based on 36 hours per week).
We also offer a set 8.3% end-of-year bonus and 8% holiday pay. For more information about our fringe benefits, please visit http://www.werkenbijvumc.nl/vumc/arbeidsvoorwaarden/ (Dutch version). The PhD position is for four years, you will start with a contract for 12 months.
For more information about the application procedure, please contact Mr. Wessel Haytink, recruiter, at +31(0)6 - 1066 7718.
Additional information
Models in this sort of context typically involve a large number of covariates p, possibly larger than the number of samples n, thus requiring specific tools. One possibility is to use random-effects models or penalized regression methods. If the focus is shifted from estimation to testing for effects of all covariates on the response, p>n problems can be handled without the need for penalization. We have extended this approach to handling multiple covariate sets. The challenge is now to develop methodsthat can find specific types of joint effects of covariate sets (molecular profiles) on the response (gene expression). Here, we will address both the testing and estimation approach, possibly in an hierarchical setup. In addition, association of these effects with clinical responses will be studied.
Interested?
For more information about the project, you may contact Dr. Renée Xavier de Menezes, via the telephonenumber +31(0)20 – 4444933; Dr. Marianne Jonker, via the telephone number +31(0)20 – 4441045; or Prof. Dr. Mark van de Wiel, via the telephone number +31(0)20 – 4442368.
Please submit your candidacy byAugust 17, 2014 at the latest with reference number D6.2014.00034WH via the following link: http://bit.ly/PhD_EandB
You may leave the list at any time by sending the command
SIGNOFF allstat
to [log in to unmask], leaving the subject line blank.
|