-
Learning the shape of protein micro-environments with a holographic convolutional neural network
Authors:
Michael N. Pun,
Andrew Ivanov,
Quinn Bellamy,
Zachary Montague,
Colin LaMont,
Philip Bradley,
Jakub Otwinowski,
Armita Nourmohammad
Abstract:
Proteins play a central role in biology from immune recognition to brain activity. While major advances in machine learning have improved our ability to predict protein structure from sequence, determining protein function from structure remains a major challenge. Here, we introduce Holographic Convolutional Neural Network (H-CNN) for proteins, which is a physically motivated machine learning appr…
▽ More
Proteins play a central role in biology from immune recognition to brain activity. While major advances in machine learning have improved our ability to predict protein structure from sequence, determining protein function from structure remains a major challenge. Here, we introduce Holographic Convolutional Neural Network (H-CNN) for proteins, which is a physically motivated machine learning approach to model amino acid preferences in protein structures. H-CNN reflects physical interactions in a protein structure and recapitulates the functional information stored in evolutionary data. H-CNN accurately predicts the impact of mutations on protein function, including stability and binding of protein complexes. Our interpretable computational model for protein structure-function maps could guide design of novel proteins with desired function.
△ Less
Submitted 5 November, 2022;
originally announced November 2022.
-
On the correspondence between thermodynamics and inference
Authors:
Colin H. LaMont,
Paul A. Wiggins
Abstract:
We expand upon a natural analogy between Bayesian statistics and statistical physics in which sample size corresponds to inverse temperature. This analogy motivates the definition of two novel statistical quantities: a learning capacity and a Gibbs entropy. The analysis of the learning capacity, corresponding to the heat capacity in thermal physics, leads to new insight into the mechanism of learn…
▽ More
We expand upon a natural analogy between Bayesian statistics and statistical physics in which sample size corresponds to inverse temperature. This analogy motivates the definition of two novel statistical quantities: a learning capacity and a Gibbs entropy. The analysis of the learning capacity, corresponding to the heat capacity in thermal physics, leads to new insight into the mechanism of learning and explains why some models have anomalously-high learning performance. We explore the properties of the learning capacity in a number of examples, including a sloppy model. Next, we propose that the Gibbs entropy provides a natural device for counting distinguishable distributions in the context of Bayesian inference. We use this device to define a generalized principle of indifference (GPI) in which every distinguishable model is assigned equal a priori probability. This principle results in a new solution to a long-standing problem in Bayesian inference: the definition of an objective or uninformative prior. A key characteristic of this new approach is that it can be applied to analyses where the model dimension is unknown and circumvents the automatic rejection of higher-dimensional models in Bayesian inference.
△ Less
Submitted 4 April, 2019; v1 submitted 5 June, 2017;
originally announced June 2017.
-
Information-based inference for singular models and finite sample sizes: A frequentist information criterion
Authors:
Colin H. LaMont,
Paul A. Wiggins
Abstract:
In the information-based paradigm of inference, model selection is performed by selecting the candidate model with the best estimated predictive performance. The success of this approach depends on the accuracy of the estimate of the predictive complexity. In the large-sample-size limit of a regular model, the predictive performance is well estimated by the Akaike Information Criterion (AIC). Howe…
▽ More
In the information-based paradigm of inference, model selection is performed by selecting the candidate model with the best estimated predictive performance. The success of this approach depends on the accuracy of the estimate of the predictive complexity. In the large-sample-size limit of a regular model, the predictive performance is well estimated by the Akaike Information Criterion (AIC). However, this approximation can either significantly under or over-estimating the complexity in a wide range of important applications where models are either non-regular or finite-sample-size corrections are significant. We introduce an improved approximation for the complexity that is used to define a new information criterion: the Frequentist Information Criterion (QIC). QIC extends the applicability of information-based inference to the finite-sample-size regime of regular models and to singular models. We demonstrate the power and the comparative advantage of QIC in a number of example analyses.
△ Less
Submitted 8 June, 2018; v1 submitted 18 June, 2015;
originally announced June 2015.
-
An objective prior that unifies objective Bayes and information-based inference
Authors:
Colin H. LaMont,
Paul A. Wiggins
Abstract:
There are three principle paradigms of statistical inference: (i) Bayesian, (ii) information-based and (iii) frequentist inference. We describe an objective prior (the weighting or $w$-prior) which unifies objective Bayes and information-based inference. The $w$-prior is chosen to make the marginal probability an unbiased estimator of the predictive performance of the model. This definition has se…
▽ More
There are three principle paradigms of statistical inference: (i) Bayesian, (ii) information-based and (iii) frequentist inference. We describe an objective prior (the weighting or $w$-prior) which unifies objective Bayes and information-based inference. The $w$-prior is chosen to make the marginal probability an unbiased estimator of the predictive performance of the model. This definition has several other natural interpretations. From the perspective of the information content of the prior, the $w$-prior is both uniformly and maximally uninformative. The $w$-prior can also be understood to result in a uniform density of distinguishable models in parameter space. Finally we demonstrate the the $w$-prior is equivalent to the Akaike Information Criterion (AIC) for regular models in the asymptotic limit. The $w$-prior appears to be generically applicable to statistical inference and is free of {\it ad hoc} regularization. The mechanism for suppressing complexity is analogous to AIC: model complexity reduces model predictivity. We expect this new objective-Bayes approach to inference to be widely-applicable to machine-learning problems including singular models.
△ Less
Submitted 24 June, 2015; v1 submitted 2 June, 2015;
originally announced June 2015.
-
The development of an information criterion for Change-Point Analysis
Authors:
Paul A. Wiggins,
Colin H. LaMont
Abstract:
Change-point analysis is a flexible and computationally tractable tool for the analysis of times series data from systems that transition between discrete states and whose observables are corrupted by noise. The change-point algorithm is used to identify the time indices (change points) at which the system transitions between these discrete states. We present a unified information-based approach t…
▽ More
Change-point analysis is a flexible and computationally tractable tool for the analysis of times series data from systems that transition between discrete states and whose observables are corrupted by noise. The change-point algorithm is used to identify the time indices (change points) at which the system transitions between these discrete states. We present a unified information-based approach to testing for the existence of change points. This new approach reconciles two previously disparate approaches to Change-Point Analysis (frequentist and information-based) for testing transitions between states. The resulting method is statistically principled, parameter and prior free and widely applicable to a wide range of change-point problems.
△ Less
Submitted 20 May, 2015;
originally announced May 2015.
-
eRHIC Design Study: An Electron-Ion Collider at BNL
Authors:
E. C. Aschenauer,
M. D. Baker,
A. Bazilevsky,
K. Boyle,
S. Belomestnykh,
I. Ben-Zvi,
S. J. Brooks,
C. Brutus,
T. Burton,
S. Fazio,
A. Fedotov,
D. Gassner,
Y. Hao,
Y. Jing,
D. Kayran,
A. Kiselev,
M. A. C. Lamont,
J. -H. Lee,
V. N. Litvinenko,
C. Liu,
T. Ludlam,
G. Mahler,
G. McIntyre,
W. Meng,
F. Meot
, et al. (22 additional authors not shown)
Abstract:
This document presents BNL's plan for an electron-ion collider, eRHIC, a major new research tool that builds on the existing RHIC facility to advance the long-term vision for Nuclear Physics to discover and understand the emergent phenomena of Quantum Chromodynamics (QCD), the fundamental theory of the strong interaction that binds the atomic nucleus. We describe the scientific requirements for su…
▽ More
This document presents BNL's plan for an electron-ion collider, eRHIC, a major new research tool that builds on the existing RHIC facility to advance the long-term vision for Nuclear Physics to discover and understand the emergent phenomena of Quantum Chromodynamics (QCD), the fundamental theory of the strong interaction that binds the atomic nucleus. We describe the scientific requirements for such a facility, following up on the community-wide 2012 white paper, 'Electron-Ion Collider: the Next QCD Frontier', and present a design concept that incorporates new, innovative accelerator techniques to provide a cost-effective upgrade of RHIC with polarized electron beams colliding with the full array of RHIC hadron beams. The new facility will deliver electron-nucleon luminosity of 10^33-10^34 cm-1sec-1 for collisions of 15.9 GeV polarized electrons on either 250 GeV polarized protons or 100 GeV/u heavy ion beams. The facility will also be capable of providing an electron beam energy of 21.2 GeV, at reduced luminosity. We discuss the on-going R&D effort to realize the project, and present key detector requirements and design ideas for an experimental program capable of making the 'golden measurements' called for in the EIC White Paper.
△ Less
Submitted 18 December, 2014; v1 submitted 4 September, 2014;
originally announced September 2014.
-
The Motion of a Pair of Charged Particles
Authors:
J. Franklin,
C. LaMont
Abstract:
We re-visit the problem of two (oppositely) charged particles interacting electromagnetically in one dimension with retarded potentials and no radiation reaction. The specific quantitative result of interest is the time it takes for the particles to fall in towards one another. Starting with the non-relativistic form, we answer this question while adding layers of complexity until we arrive at the…
▽ More
We re-visit the problem of two (oppositely) charged particles interacting electromagnetically in one dimension with retarded potentials and no radiation reaction. The specific quantitative result of interest is the time it takes for the particles to fall in towards one another. Starting with the non-relativistic form, we answer this question while adding layers of complexity until we arrive at the full relativistic delay differential equation that governs this problem. That case can be solved using the Synge method, which we describe and discuss.
△ Less
Submitted 17 October, 2013;
originally announced October 2013.
