| KEYWORD |
Area Engineering
Benchmarking ancestral protein reconstruction methods using lattice models
keywords BIOLOGICAL EVOLUTION, STATISTICAL INFERENCE, STATISTICAL PHYSICS
Reference persons PIERRE BARRAT-CHARLAIX, ANDREA PAGNANI
Research Groups AA - Statistical Physics and Interdisciplinary Applications
Thesis type MASTER THESIS
Description Proteins are highly complex molecules, essential to all living cells. In the course of biological evolution, they diversify through mutation while natural selection preserves their function. As a result, protein families contain thousands of sequences that are highly variable and yet maintain similar functions. On the other hand, even a few mutations can destabilize a protein and destroy its function.
The goal of Ancestral Sequence Reconstruction (ASR) consists in inferring likely sequences for extinct proteins that are ancestral to present ones. For this, it is necessary to combine a model of biological evolution with inference techniques on trees. ASR is a useful technique, but it is hard to measure its accuracy and potential biases since real ancestral proteins are almost always unknown [1].
The goal of the thesis will be to develop an evolution model for Lattice Proteins [2,3], a simplified model of actual proteins, and to use it to benchmark ancestral reconstruction methods. In particular, it will be used to investigate the claim that ancestral proteins are more thermostable than extent ones.
References:
[1] Lucas C Wheeler, Shion A Lim, Susan Marqusee, Michael J Harms, The thermostability and specificity of ancient proteins, Current Opinion in Structural Biology, 2016
[2] Jesse D. Bloom, Jonathan J. Silberg, Claus O. Wilke, D. Allan Drummond, Christoph Adami, and Frances H. Arnold, Thermodynamic prediction of protein neutrality, PNAS, 2005
[3] Emanuele Loffredo, Elisabetta Vesconi, Rostam Razban, Orit Peleg, Eugene Shakhnovich, Simona Cocco, Rémi Monasson, Evolutionary Dynamics of a Lattice Dimer: a Toy Model for Stability vs. Affinity Trade-offs in Proteins, Journal of Physics A, 2023
Required skills - Background in statistical physics / computational physics
- Knowledge of the Julia programming language is a plus but not required.
- Familiarity with the topics of biological evolution and/or proteins is not required.
Deadline 19/07/2025
PROPONI LA TUA CANDIDATURA