In this course, we collectively work on student’s real (post-QE) projects and data, to derive, improve, or critically scrutinize models or data analysis techniques, to suggest alternatives, or possibly discover new interesting structures in the data.
Specifically, the aims of this course are as follows:
-) Learn to choose an appropriate model or data analysis method for the project.
-) Learn, try out and teach other students about new data analysis approaches.
-) Deal with limitations of real data (data size, normalizations, lack of calibration etc.).
-) Learn to communicate with people with whom you share data-analysis background, but not the biological topic or system; much in contrast to the typical case.
-) Learn structured communication. For their project, the student maintains a written record (see below for details): (i) a 2-page abstract of the project focusing on data-analysis issues, (ii) the initial presentation, (iii) key questions from the group and student’s responses (mimicking referee responses), (iv) key open data analysis issues identified by the student and suggestions by the group to explore, for all 4 cycles, (v) the final presentation.
Since this course deals with real-world problems and is not a theory course, there are typically no “absolutely correct” solutions. Rather, the point is to give suggestions and try out and evaluate alternative approaches. The role of the Instructors is to evaluate the suggestions from the group and the student to identify and make precise the tasks that the student should do as a homework and report on until the next cycle.
Target group: Students with strong quantitative background working on biological problems. This course is targeted specifically to support students with math / coding / physics / engineering background working on life-science problems, typically in life-science or interdisciplinary groups. Students with life-science backgrounds can register only if they have sufficient knowledge of math and coding (see Prerequisites).
Prerequisites: Students should be well-versed in coding and some data analysis, being able to write their own numerical code, simulation, or scripts and not just use existing packages. The required level of background is the same as for taking the DSSC Track Core course. This course does not teach basic modeling or basic coding to biology students!
Evaluation: Active participation (100%).
Teaching format: Every week, groups comprising up to 6 students, the Instructor + TA, focus on two students’ projects. With 6 students, it takes 3 weeks to cycle through each student’s project, and the class will go through 4 such cycles. In the first cycle, each student presents their project, focusing less on the biological issues and more on the data/data analysis / modeling / aim side of the project. For the initial presentation, the student also prepares a short, 2-page summary of the relevant information of the project (following a structure prescribed by the Instructor) that will serve throughout the course as an “abstract” for that project. After the presentation, the group prepares a list of ~5 key questions to be clarified or quantified by the student until next meeting and gives general first-impression feedback. In the second cycle, the student prepares written responses to the key questions and presents them. The student also presents key open data analysis questions or challenges, from their own perspective. The group suggests new or alternative analyses to the student to support the of student’s project aim and address these challenges. In the third cycle, the student presents the new analyses and gets feedback from the group; the point is to also be pedagogic: so, if the student decided to learn about and use, e.g., multi-dimensional scaling, they should be able to explain to the group the basic idea of such a method. This provides a structured learning experience to the group. The group then either makes one new round of recommendations or alternatively, suggests extra directions for data exploration that may uncover interesting features of the data. In the fourth cycle, the students present the final stage of their project in a broader context (e.g., research talk) discuss plans for the future. Feedback is given by the group on the complete presentation, that is, on the broad scientific question and conclusions, rather than focusing on solely data analysis methods.
ECTS: 3 Year: 2021
Track segment(s):
BIO-QUANT Biology - Quantitative and Computational Methods in Biology
Teacher(s):
Gasper Tkacik
Teaching assistant(s):
If you want to enroll to this course, please click: REGISTER
- Trainer/in: Gasper Tkacik
- Teaching Assistant: Wiktor Mlynarski