My personal and professional adventure into Science

Exam in Introduction to Bioinformatics

Last night I most likely had a student or two that had difficulties sleeping, hopefully many was looking forward to this day, as today was the exam day for the course I am co-responsible for: “Introduction to Bioinformatics”. 

109 students was this year signed up for the course and 99 had registered for the exam. The examination setup is a 4 hour written exam, where the students will touch upon the tools and subjects that they have been taught during the course. This is among others databases like GenBank , UniProt and PDB, Pairwise and multiple alignment, Phylogenetic trees, BLAST, LOGO-plots, Weight-matrices, Prediction methods and other interesting and fun stuff 🙂

Now the exam is closed, 89 students handed in, and I now have three weeks to correct all the exam sets. The students might have been sweating today, but now it’s my turn to sweat while correcting them all. I’m looking forward to see how they all performed 🙂

 

First danish Coursera course – Learn your Computational Molecular Evolution for free

Here at CBS we are very proud to announce that the first danish Coursera course – Computational Molecular Evolution – taught by Professor Anders Gorm Pedersen will start Monday June 24 (www.coursera.org/course/molevol).

Computational Molecular Evolution

Coursera is a platform for Massive Open Online Courses (MOOCs) offered to the world for free (at coursera.org). It was founded last year by two visionary professors from Stanford University, Andrew Ng and Daphne Koller, under the motto: Education should not be a privilege – but a human right. Andrew and Daphne have not been aiming for any kind of education, they have been working hard to secure that only the best universities, offering the best courses by their best instructors, and using the most efficient pedagogical tools are available on Coursera.

Coursera has since at many occasions been referred to as a revolution in higher education, and has in only a little more than a year grown from 0 to almost 4 million students (courserians). Not surprisingly, professor Andrew Ng and Daphne Koller are now among the Time’s top 100 most influential people in the world (time100.time.com/2013/04/18/time-100/slide/andrew-ng-and-daphne-koller).

Computational Molecular Evolution has more than 13,000 students enrolled, and will run for 6 weeks for the first time this summer. It is a transformed version of a course that Professor Anders Gorm Pedersen is teaching at DTU, and the product now available at Coursera is the result of months of hard work by a range of people from both DTU and Coursera.

This course is about molecular evolution – the evolution of DNA, RNA, and protein molecules. The focus is on computational methods for inferring phylogenetic trees from sequence data, and the course will give an introduction to the fundamental theory and algorithms, while also giving the student hands-on experience with some widely used software tools. Since evolutionary theory is the conceptual foundation of biology (in the words of Theodosius Dobzhansky: “Nothing in biology makes sense except in the light of evolution”), what you learn on this course will be relevant for any project you will ever do inside the life sciences. A phylogenetic tree will almost always help you think more clearly about your biological problem.

A special emphasis is put on methods that employ explicit models of the evolutionary process (maximum likelihood and Bayesian approaches), and we will explore the role of statistical modeling in molecular evolution, and in science more generally. A mathematical (statistical) model of a biological system can be considered to be a stringently phrased hypothesis about that system, and this way of thinking about models will often be helpful. In addition to model-based methods, you will also learn about other approaches, such as those based on parsimony and genetic distance (e.g., neighbor joining).

Often, the evolutionary tree is the result we are interested in – knowing how a set of sequences (or organisms) are related can provide us with important information about the biological problem we are  investigating. For instance, knowing which organisms are most closely related to a newly identified, uncharacterized, pathogenic bacterium will allow you to infer many aspects of its lifestyle, thereby giving you important clues about how to fight it. In other cases, however, inferring the structure of the tree is not the goal: for instance, our main focus may instead be the detection of positions in a protein undergoing positive selection (indicating adaptation) or negative selection (indicating conserved functional importance). However, even in these cases, the underlying phylogenetic tree will be an important part of our hypothesis about (model of) how the proteins have been evolving, and will help in getting the correct answer.

Although the study of molecular evolution does require a certain level of mathematical understanding, this course has been designed to be accessible also for students with limited computational background (e.g., students of biology)

The team is very excited about the launch, and hope that interested students around the world will enjoy it.

Computational Molecular Evolution

Media coverage:

Text in danish:Dr.dk: 14.000 studerende fra hele verden følger DTU-kursus
http://karriere.jobfinder.dk/artikel/dtu-professor-foerste-dansker-paa-gratis-undervisningsportal-272

PhD Summer School on Machine Learning 2013 at DTU Compute

During my PhD I have used Machine Learning to train Neural Networks. It is an excellent tool to find patterns in data and train models to predict on unknown data. In august DTU compute is offering a Summer School for PhD’s on Machine Learning. I haven’t attended the course myself, you can see the invitation below.

 

********************************************************************************

PhD Summer School on Machine Learning
(www.imm.dtu.dk/courses/02901)

August 12 – 16, 2013 (both days included)
DTU Compute, Technical University of Denmark,

********************************************************************************
We hereby invite you to participate in the PhD Summer School on Machine Learning that will take place at the Technical University of Denmark.

The course will consist of lectures given by invited speakers with expertise in machine learning as well as members of the research groups at DTU Compute. The course will cover key topics in machine learning including Bayesian parametric and non-parametric inference, optimization, low rank matrix factorization and kernel methods. As preparation for the course we recommend reading the book by Christopher M. Bishop “Pattern Recognition and Machine Learning”, Springer 2006 (chapter 1-6 and 9). The course requires basic Matlab programming skills.

Confirmed invited Speakers:

Joaquin Quiñonero Candela, Facebook

Ryota Tomioka, University of Tokyo

Further information regarding the summer school can be found on our webpage:
www.imm.dtu.dk/courses/02901

For registration, please contact: Marian Solrun Adler masad@dtu.dk or Morten Mørup mmor@dtu.dk

We hope to see you at the campus of the Technical University of Denmark!

The organizers,
Morten Mørup, Lars Kai Hansen, Mikkel N. Schmidt and Ole Winther

%d bloggers like this: