Bayesian protein superposition using Hamiltonian Monte Carlo
Research output: Chapter in Book/Report/Conference proceeding › Article in proceedings › Research › peer-review
Standard
Bayesian protein superposition using Hamiltonian Monte Carlo. / Moreta, Lys Sanz; Al-Sibahi, Ahmad Salim; Hamelryck, Thomas.
Proceedings - IEEE 20th International Conference on Bioinformatics and Bioengineering, BIBE 2020. IEEE, 2020. p. 1-11 9288019.Research output: Chapter in Book/Report/Conference proceeding › Article in proceedings › Research › peer-review
Harvard
APA
Vancouver
Author
Bibtex
}
RIS
TY - GEN
T1 - Bayesian protein superposition using Hamiltonian Monte Carlo
AU - Moreta, Lys Sanz
AU - Al-Sibahi, Ahmad Salim
AU - Hamelryck, Thomas
PY - 2020/10
Y1 - 2020/10
N2 - Optimally superimposing protein structures is essential to study their structure, function, dynamics and evolution. We present THESEUS NUTS (No U-Turn Sampler), a Bayesian version of the THESEUS model [1] -[3] which relies on maximum likelihood estimation. The probabilistic model interprets each protein as a rotated and translated noisy observation of a latent mean structure. Unlike conventional methods [4], THESEUS takes into account the differences in correlations between the atoms in the structure. This paper extends the previous THESEUS MAP (Maximum A Posteriori) model, [5] to full Bayesian inference by making use of the iterative NUTS [6], a Hamiltonian Monte Carlo method. The model delivers consistent results and is computationally efficient thanks to its implementation in the probabilistic programming language NumpPyro [7], [8] which in turn relies upon JAX [9], a system for high-performance machine learning.
AB - Optimally superimposing protein structures is essential to study their structure, function, dynamics and evolution. We present THESEUS NUTS (No U-Turn Sampler), a Bayesian version of the THESEUS model [1] -[3] which relies on maximum likelihood estimation. The probabilistic model interprets each protein as a rotated and translated noisy observation of a latent mean structure. Unlike conventional methods [4], THESEUS takes into account the differences in correlations between the atoms in the structure. This paper extends the previous THESEUS MAP (Maximum A Posteriori) model, [5] to full Bayesian inference by making use of the iterative NUTS [6], a Hamiltonian Monte Carlo method. The model delivers consistent results and is computationally efficient thanks to its implementation in the probabilistic programming language NumpPyro [7], [8] which in turn relies upon JAX [9], a system for high-performance machine learning.
KW - Bayesian modelling
KW - Hamiltonian Monte Carlo
KW - NUTS
KW - probabilistic programming
KW - protein structure superposition
KW - protein superposition
UR - http://www.scopus.com/inward/record.url?scp=85099556496&partnerID=8YFLogxK
U2 - 10.1109/BIBE50027.2020.00009
DO - 10.1109/BIBE50027.2020.00009
M3 - Article in proceedings
AN - SCOPUS:85099556496
SP - 1
EP - 11
BT - Proceedings - IEEE 20th International Conference on Bioinformatics and Bioengineering, BIBE 2020
PB - IEEE
T2 - 20th IEEE International Conference on Bioinformatics and Bioengineering, BIBE 2020
Y2 - 26 October 2020 through 28 October 2020
ER -
ID: 255836764