I did not see any other thread about this yet. The webforum seems a little quiet ... to word this nicely.

Anyway, exam questions today - 28.01.2026

Note that 40 questions came. I only remember a tiny bit.

The first part, say about 16 questions or so, perhaps 13, came about python.

This covered the basic stuff mostly. For instance what will this do in Python:

'ab'[::-1]

Also, some basic questions ... int float difference, String, which basic
operations on String (e. g. will + work will slicing work and so forth].

Also if-like conditional, assignment, flow-logic in python.

Interestingly some questions about indentation. One was whether indentation
is used to denote code blocks.

Hmm logical operators too. But basically basic python questions.

Machine Learning was a bit harder. Here I recommend solid math questions. You may have to also understand matrices a little, e. g. whether you can inverse the following matric, and also some basic formula where you need to figure out how a matrix changes, if at all. This looks complicated but you can simplify it. But you need to understand matrices.

Some questions about unsupervised learning and model building e. g. test data training data, difference, which to use. Also one question was about what to pick.

From simulation a few questions about protein structure e. g. alpha helix residues where are they? And CATH top domain classification, alpha-domains, beta-domains or alpha-beta domains. . Also you should know what AlphaFold does in principle but also other software; one question came here, from structure to sequence (I only knew from sequence to structure, that question surprised me, so make sure you know software for both forward and reverse steps).

From imaging one question was about coordinate systems. And the other was about mathematica or so, probably done in the exercise (I didn't do the exercise but it is recommended to do both, so it may be better to actually indeed do both at the same time, I think you will probably learn most. I only took the VO as I didn't have time for exercises this semester.)

Edit: Sorry the software was called ProteinMPNN. Never heard it before, but this thing can evidently predict a sequence, from a structure, with some probability. Not sure how this works but that's quite interesting; I haven't heard from this before.

Ah yes, first and second exam will be open book so you can take a notebook with you. So the python questions should be super-simple even for people who don't know python that well.

The Machine Learning questions required a bit more preparation in my opinion. I think it is beneficial if you go through the ~5 main algorithms and understand them in principle or so. One exam question mentioned an algorithm and I actually didn't study that one in detail so I had no idea what to select there.