- Syllabus
- Lecture 1: Introduction
- Lecture 2: Kmeans and mixtures, with applications to torsion angles in flat torii
- Lecture 3: From Darwin to AlphaFold: MSA, DCA, attention mechanisms, and structure prediction
- Lecture 4: Inside AlphaFold
- Lecture 5: Exploring high dimensional spaces, molecular kinematics, loop sampling
- Lecture 6: Nearest neighbors in high dimensional spaces: algorithms and significance
- Lecture 7: Generative models via transport equations, applications to molecular conformations
- (Extra slides) Structural alignments and analysis on static structures
- (Extra slides) Motions and energies: normal modes, (free) energies, ICA, tICA
The exam will take place on April the 9th, following the protocol:
- 03/01: release of the papers for the projects. Students will
have one week to form pairs and select the projects--first come first serve
or mutual agreement.
- 04/09: 9am-10am or 2pm-3pm: paper and pencil written exam. class notes allowed, but no electronic device.
- 04/09: 10:30-5pm: project presentations.
- Students will present in tandem one the research paper below.
Bonus point will be awarded for two intiatives:
(i) computer experiments aiming
at reproducing the results; (ii) a well-founded strategy to improve on the paper and its limitations.
- For a given presentation: all students will attend it, *except* those working on the same project, who did not present yet.
The precise schedule for this second part will be provided later.
- (Master/PhD) Sampling conformations in the generator/operator matching framework
Contact: Frederic.Cazals@inria.fr
Class 2024-25 -- with slides