Lecture and Exercise Protein Prediction II for Computational Biologists WiSe 2020/21


 

Type:

Lecture (4 SWS) + Exercise (2 SWS)

Ects:

8.0

Lecturer

Burkhard Rost  (lecture)

Michael Heinzinger (exercise)

Rotation:

Tuesday, 12:30 - 14:00
Thursday, 12:30 - 14:00
Thursday, (exercise), 14:00 - 15:00

Place:

Lecture:  virtually

Exercise: virtually

Exam: tba

Language:

English

Announcements:

The latest news are always on top!

First lecture on Tuesday, Nov 10. 12:15am CEST. 
Details in zoom-slides

Our lecture and exercise will be completely virtually.

Exam

The link for taking part in online proctoring is sent via mail. If you do not have it by 01.02.2021, 10:00 PM CET, please contact pp1ex@rostlab.org. 

Piazza

piazza.com/tum.de/spring2021/in2230in2291

 

Content

This lecture continues the 'Protein Prediction I' lecture. The attendance of this lecture however is not a mandatory requirement for 'Protein Prediction II'. Topics will include (but not be limited to):

  • Predicting protein function using sequence: sequence alignments, multiple sequence alignments, motifs, domain assignment, annotation transfer by homology, ab initio predictions.
  • Predicting protein function using structure: structural alignments, structural motifs, annotation transfer via structure similarity.
  • From structure prediction to function prediction: comparative modeling; prediction of: secondary structure, hydrophilicity profiles, solvent accessibility, transmembrane segments, disordered regions, contact maps, functional residues; template free modeling.
  • Machine learning.

Slides

  1. Zoom-Slides  Lecture Plan
  2. Lecture 1: Intro Nov 10th - Live Session
  3. Lecture 2: Intro to Protein Function Nov 11th - Video
  4. Lecture 3: Predict Localization  Nov 19th  - Video
  5. Lecture 4: Localization1 Nov 24th - Live on Zoom (12:30pm)
  6. Lecture 5: Localization2 Nov 29th - Live on Zoom (12:30pm) Video
  7. Lecture 6: Localization3 Dec 1st - Video
  8. Lecture 7: Protein Protein Interaction Sites Dec 8th - Live on Zoom (12:30pm)
  9. Lecture 8: Protein Protein Interaction Pairs Dec 10th - Video
  10. Lecture 9: Binding Site & Ligand Interaction Prediction Dec 15th - Video
  11. Lecture 10: Representation Learning Jan 12th - Live on Zoom (start 12:30pm CET)
  12. Lecture 11: Deep Learning Jan 14th - Live on Zoom (start 12:30pm CET)
  13. Lecture 12: SAV effect 1 Jan 19th - Live on Zoom  (start 12:30pm CET) Video1   Video2 Website Issue. Slides temporarily hosted on LRZ Sync+Share
  14. Wrap Up: mock exam - start 12:30pm CET
  15. Wrap Up: Live on Zoom (start 12:30pm CET)

Exercise

Slides

To successfully complete the exercise each group needs to fulfill the following requirements:

  • Give a presentation regarding their project that focuses either on the biological background, current state-of-the-art methods, or the provided data set.
  • Give a final presentation about their approach; e.g. architecture and performance.
  • Actively participate in the exercise. This primarily means attending all students' presentations and participating in the corresponding Q&A sessions.
  • Submit a working solution at the end of the semester. It should be able to make informed predictions based on new data and perform better than a random or very naive solution.

Presentations in general:

  • Attendance is mandatory for all students (not just the groups presenting) whenever there is a presentation by students.
  • For sessions without students' presentations, attendance is not mandatory.

Working in a group:

  • The workload within each group should be evenly split.
  • Not every student has to speak during their group's presentation (e.g. some could work on the presentations, the rest on the programming task).
  • If students in your group refuse to "carry their weight" or you are the only one doing any work, feel free to contact us (better early than late).