-
Binding prediction of multi-domain cellulases with a dual-CNN
Authors:
Kay S. Schaller,
Jeppe Kari,
Kim Borch,
Günther H. J. Peters,
Peter Westh
Abstract:
Cellulases hold great promise for the production of biofuels and biochemicals. However, they are modular enzymes acting on a complex heterogeneous substrate. Because of this complexity, the computational prediction of their catalytic properties remains scarce, which restricts both enzyme discovery and enzyme design. Here, we present a dual-input convolutional neural network to predict the binding…
▽ More
Cellulases hold great promise for the production of biofuels and biochemicals. However, they are modular enzymes acting on a complex heterogeneous substrate. Because of this complexity, the computational prediction of their catalytic properties remains scarce, which restricts both enzyme discovery and enzyme design. Here, we present a dual-input convolutional neural network to predict the binding of multi-domain enzymes. This regression model outperformed previous molecular dynamics-based methods for binding prediction for cellulases in a fraction of the time. Also, we show that when changed to a classification problem, the same network can be back-propagated to suggest mutations to improve enzyme binding. A similar approach could increase our understanding of the structure-activity relationship of enzymes, and suggest new promising mutations for enzyme design using explainable artificial intelligence.
△ Less
Submitted 6 July, 2022;
originally announced July 2022.
-
The SLAC T-510 experiment for radio emission from particle showers: detailed simulation study and interpretation
Authors:
K. Bechtol,
K. Belov,
K. Borch,
P. Chen,
J. Clem,
P. Gorham,
C. Hast,
T. Huege,
R. Hyneman,
K. Jobe,
K. Kuwatani,
J. Lam,
T. C. Liu,
K. Mulrey,
J. Nam,
C. Naudet,
R. J. Nichol,
C. Paciaroni,
B. F. Rauch,
A. Romero-Wolf,
B. Rotter,
D. Saltzberg,
H. Schoorlemmer,
D. Seckel,
B. Strutt
, et al. (4 additional authors not shown)
Abstract:
Over the last several decades, radio detection of air showers has been widely used to detect ultra-high-energy cosmic rays. We developed an experiment under controlled laboratory conditions at SLAC with which we measured the radio-frequency radiation from a charged particle shower produced by bunches of electrons as primaries with known energy. The shower took place in a target made of High Densit…
▽ More
Over the last several decades, radio detection of air showers has been widely used to detect ultra-high-energy cosmic rays. We developed an experiment under controlled laboratory conditions at SLAC with which we measured the radio-frequency radiation from a charged particle shower produced by bunches of electrons as primaries with known energy. The shower took place in a target made of High Density Polyethylene located in a strong magnetic field. The experiment was designed so that Askaryan and magnetically-induced components of the radio emission could be measured independently. At the same time, we performed a detailed simulation of this experiment to predict the radio signal using two microscopic formalisms, endpoint and ZHS. In this paper, we present the simulation scheme and make a comparison with data characteristics such as linearity with magnetic field and amplitude. The simulations agree with the measurements within uncertainties and present a good description of the data. In particular, reflections within the target that accounted for the largest systematic uncertainties are addressed. The prediction of the amplitude of Askaryan emission agrees with measurements to within 5% for the endpoint formalism and 11% for the ZHS formalism. The amplitudes of magnetically-induced emission agree to within 5% for the endpoint formalism and less than 1% for the ZHS formalism. The agreement of the absolute scale of emission gives confidence in state-of-the-art air shower simulations which are based on the applied formalisms.
△ Less
Submitted 3 March, 2022; v1 submitted 8 November, 2021;
originally announced November 2021.
-
Accelerator measurements of magnetically-induced radio emission from particle cascades with applications to cosmic-ray air showers
Authors:
K. Belov,
K. Mulrey,
A. Romero-Wolf,
S. A. Wissel,
A. Zilles,
K. Bechtol,
K. Borch,
P. Chen,
J. Clem,
P. W. Gorham,
C. Hast,
T. Huege,
R. Hyneman,
K. Jobe,
K. Kuwatani,
J. Lam,
T. Liu,
J. Nam,
C. Naudet,
R. Nichol,
B. F. Rauch,
B. Rotter,
D. Saltzberg,
H. Schoorlemmer,
D. Seckel
, et al. (3 additional authors not shown)
Abstract:
For fifty years, cosmic-ray air showers have been detected by their radio emission. We present the first laboratory measurements that validate electrodynamics simulations used in air shower modeling. An experiment at SLAC provides a beam test of radio-frequency (RF) radiation from charged particle cascades in the presence of a magnetic field, a model system of a cosmic-ray air shower. This experim…
▽ More
For fifty years, cosmic-ray air showers have been detected by their radio emission. We present the first laboratory measurements that validate electrodynamics simulations used in air shower modeling. An experiment at SLAC provides a beam test of radio-frequency (RF) radiation from charged particle cascades in the presence of a magnetic field, a model system of a cosmic-ray air shower. This experiment provides a suite of controlled laboratory measurements to compare to particle-level simulations of RF emission, which are relied upon in ultra-high-energy cosmic-ray air shower detection. We compare simulations to data for intensity, linearity with magnetic field, angular distribution, polarization, and spectral content. In particular, we confirm modern predictions that the magnetically induced emission in a dielectric forms a cone that peaks at the Cherenkov angle and show that the simulations reproduce the data within systematic uncertainties.
△ Less
Submitted 22 April, 2016; v1 submitted 26 July, 2015;
originally announced July 2015.
