3. January, 2019
Report from exchange to École Polytechnique Fédérale de Lausanne (EPFL)
Thanks to NORBIS I got the unique opportunity to experience a different laboratory in Switzerland for 6 months. First, upon arrival in Switzerland it took some time to organize everything before being able to start experiments. However, as soon as everything was in place the pace at which my host laboratory worked was overwhelming and allowed me to finish many important experiments in a rather short time.
I learned different new techniques such as QTL mapping and usage of big datasets. In particular, the usage of a genetic reference population allowed me to identify new potential mechanisms driven by my gene of interest. In summary, both my research project and me personally benefited tremendously from this research stay.
Most importantly, I was very lucky to meet outstanding colleagues with whom I spent long nights in the laboratory learning various things, but additionally we also shared beautiful moments in the swiss mountains.
In general, I would recommend planning more time than anticipated as settling in a new laboratory takes time. However, once the first hurdle is overcome it is an unforgettable time that everyone should experience at some point.
17. December, 2018
Developing new Bayesian models in London
I am a PhD student at the Department of Biostatistics of the University of Oslo. I had stayed in London for four months funded by the NORBIS, as a visiting PhD student with the group of Dr. Alex Lewin, who is Associate Professor in Biostatistics at the Department of Medical Statistics at the London School of Hygiene & Tropical Medicine.
London School of Hygiene & Tropical Medicine, London, UK
I was collaborating with Dr. Alex Lewin and Dr. Marco Banterle on new Bayesian models for drug sensitivity prediction and integration of multi-omics data. The developed computational tools for the analysis of these data can consider the intrinsic relationships between the various omics data sources and also between different anti-cancer drugs, and generate new biological knowledge by helping us to identify which omics data sources and which individual features are most predictive for the sensitivity of which (classes of) drugs.
We have adapted a promising modelling approach, that was previously developed in Dr. Lewin’s group, to our situation. We have established a new Bayesian model framework for drug sensitivity prediction and drug targets identification. The new Bayesian model uses Seemly Unrelated Regressions for estimating a large covariance matrix efficiently, a spike-and-slab prior for selecting sparsely relevant molecular features, and a Markov random field prior for capturing the drug-drug similarity and related targeted genes/pathways. During my stay in the UK, I also had a good opportunity to visit the MRC Biostatistics Unit at the University of Cambridge and have nice discussions with some researchers there. Although I am now back in Oslo, our collaboration on this project is continuing and will result in a joint publication.
This year London had a very good summer, a lot of sunny days rather than mostly rainy days. The best relaxing ways for me were taking a stroll along the Thames and enjoying the Hampstead Heath walking. But it was extremely hot some days in July, especially in many buildings and on the underground trains (the “Tube”) without air conditioners. In General, I had an enjoyable summer visit in London. However, I don’t highly recommend others for such short international exchange during summer. It might be better to avoid the holiday season,
so that you could have more opportunities to discuss with your collaborators.
In addition, since London is a super-rich city and one of the world’s largest trading centres, it is difficult to find a not so expensive accommodation for a short stay. I regret not to spend more time on looking for one fixed accommodation rather than living in three places during four months.
Finally, I would like to thank NORBIS for the travel grant, and the collaborators in London very much.
Tower Bridge, London, UK
20. August, 2018
By: Anna-Simone Josefine Frank
I’m currently a PhD student at the School of Pharmacy (UiO) with a background in mathematics. A travel grant from NORBIS, for which I’m grateful, enabled me to spend six months (January to June 2018) as Visiting Fellow at Cornell University in New York, USA. My main affiliation at Cornell was with the Department of Biological Statistics and Computational Biology (BSCB).
Picture: On campus in early 2018 (Winter at Cornell was harsh)
My PHD project and purpose of my stay
My PhD research focuses on the use of thyroid hormone replacement therapy (THRT) during pregnancy. The goal is to investigate whether THRT has effect on immediate birth outcomes. The project relies on the use of statistical methodologies and analysis tools. Hence the goal of my visit at Cornell was to spend time in a quantitative research environment renowned for the development of novel statistical methodologies.
The Department and Collaboration
I joined the research group of Professor Matteson early January, and felt very welcome by him, students, faculty members and administrative staff. I was immediately offered office space and the necessary research facilities. Professor Matteson and I agreed on regular meetings. These meetings created a forum for discussion, where I have had to answer questions that have contributed to deepen my own understanding of the project. Most challenging however, was explaining to a statistician with no background in pharmacoepidemiology what my research is about.
Outcomes of Visit
Our first collaborative project aimed to classify women according to similar patterns of medication use, and to compare these patterns across different data sources. We applied Group-based trajectory models (GBTM) to the data and identified four disjoint groups of adherence patterns of medication use. The results were summarized in a manuscript, entitled ‘Group-based trajectory models to determine patterns from different data sources on maternal use of thyroid hormone replacement therapy’. This manuscript, submitted whiles at Cornell, is currently under review.
As an extension of this project, we have started work on quantifying the effect of THRT use on immediate pregnancy outcomes. Collaboration on this project will continue beyond my research visit.
Picture: Cornell Tower (left) and Flower Garden on campus in summer (right)
Academic Seminars and Workshops
In addition to research, I attended weekly seminars at the department of statistics as well as took part in workshops on statistical methods, which was organized by the Cornell Statistical Consulting Unit (CSCU). Graduate students at the department usually present their latest research results during a bi-weekly seminar. I had the opportunity of presenting my research project during one of the seminar meetings in March.
Besides academia, I supported the Cornell University hockey team (the Big Red), during their their Ivy league hockey tournament. In April, I also participated as a panelist to share my career path experience with undergraduate students at Cornell University. In May, I got accepted to participate in a “Julie Tumbles Leadership retreat” workshop tailored specifically to young, female researchers and organized by CORNELL’S GRADUATE & PROFESSIONAL WOMEN’S NETWORK (GPWOMEN). The workshop discussed the obstacles women face during their career and how to deal with them.
Picture: Hockey trophies, tickets (up) and Lynah Rink (down)
In total, my six months stay abroad at the department of BSCB was very stimulating experience both personally and academically. So far it was the best and most rewarding experience during my PhD, and I’m grateful to NORBIS for making this possible. I can strongly recommend every PhD student to spend some time abroad.