Work

Auf Maschinellem Lernen gestützte Methoden für das de novo design von Proteinen mittels Netzwerken zur Strukturvorhersage

Year: 2025

Type: dissertation

Abstract: This work proposes two new de novo protein design strategies: Relaxed sequence optimization (RSO) and the af2cycler. With RSO, we build on the commonly used hallucination approach and introduce a relaxed sequence space, representing a superposition of all possible sequences at every residue position. The second strategy, the af2cycler, is a method to increase the in silico and experimental success rate of designs from other methods like Chroma, by employing iterative refinement using Alphafold2.

Source: mediaTUM – the media and publications repository of the Technical University Munich (Technical University Munich)

Author: Frank, Christopher Josef

Language: English

Topic: Protein Structure and Dynamics

Subfield: Molecular Biology

Field: Biochemistry, Genetics and Molecular Biology

Domain: Life Sciences

Open Access status: green

OpenAlex ID: W7117085716

Cites:

Cited by:

Related to: 10