-
Neurosymbolic AI for Reasoning on Biomedical Knowledge Graphs
Authors:
Lauren Nicole DeLong,
Ramon Fernández Mir,
Zonglin Ji,
Fiona Niamh Coulter Smith,
Jacques D. Fleuriot
Abstract:
Biomedical datasets are often modeled as knowledge graphs (KGs) because they capture the multi-relational, heterogeneous, and dynamic natures of biomedical systems. KG completion (KGC), can, therefore, help researchers make predictions to inform tasks like drug repositioning. While previous approaches for KGC were either rule-based or embedding-based, hybrid approaches based on neurosymbolic artif…
▽ More
Biomedical datasets are often modeled as knowledge graphs (KGs) because they capture the multi-relational, heterogeneous, and dynamic natures of biomedical systems. KG completion (KGC), can, therefore, help researchers make predictions to inform tasks like drug repositioning. While previous approaches for KGC were either rule-based or embedding-based, hybrid approaches based on neurosymbolic artificial intelligence are becoming more popular. Many of these methods possess unique characteristics which make them even better suited toward biomedical challenges. Here, we survey such approaches with an emphasis on their utilities and prospective benefits for biomedicine.
△ Less
Submitted 17 July, 2023;
originally announced July 2023.
-
Optimizing accuracy and efficacy in data-driven materials discovery for the solar production of hydrogen
Authors:
Yihuang Xiong,
Quinn T. Campbell,
Julian Fanghanel,
Catherine K. Badding,
Huaiyu Wang,
Nicole E. Kirchner-Hall,
Monica J. Theibault,
Iurii Timrov,
Jared S. Mondschein,
Kriti Seth,
Rebecca Katz,
Andres Molina Villarino,
Betül Pamuk,
Megan E. Penrod,
Mohammed M. Khan,
Tiffany Rivera,
Nathan C. Smith,
Xavier Quintana,
Paul Orbe,
Craig J. Fennie,
Senorpe Asem-Hiablie,
James L. Young,
Todd G. Deutsch,
Matteo Cococcioni,
Venkatraman Gopalan
, et al. (3 additional authors not shown)
Abstract:
The production of hydrogen fuels, via water splitting, is of practical relevance for meeting global energy needs and mitigating the environmental consequences of fossil-fuel-based transportation. Water photoelectrolysis has been proposed as a viable approach for generating hydrogen, provided that stable and inexpensive photocatalysts with conversion efficiencies over 10% can be discovered, synthes…
▽ More
The production of hydrogen fuels, via water splitting, is of practical relevance for meeting global energy needs and mitigating the environmental consequences of fossil-fuel-based transportation. Water photoelectrolysis has been proposed as a viable approach for generating hydrogen, provided that stable and inexpensive photocatalysts with conversion efficiencies over 10% can be discovered, synthesized at scale, and successfully deployed (Pinaud et al., Energy Environ. Sci., 2013, 6, 1983). While a number of first-principles studies have focused on the data-driven discovery of photocatalysts, in the absence of systematic experimental validation, the success rate of these predictions may be limited. We address this problem by develo** a screening procedure with co-validation between experiment and theory to expedite the synthesis, characterization, and testing of the computationally predicted, most desirable materials. Starting with 70,150 compounds in the Materials Project database, the proposed protocol yielded 71 candidate photocatalysts, 11 of which were synthesized as single-phase materials. Experiments confirmed hydrogen generation and favorable band alignment for 6 of the 11 compounds, with the most promising ones belonging to the families of alkali and alkaline-earth indates and orthoplumbates. This study shows the accuracy of a nonempirical, Hubbard-corrected density-functional theory method to predict band gaps and band offsets at a fraction of the computational cost of hybrid functionals, and outlines an effective strategy to identify photocatalysts for solar hydrogen generation.
△ Less
Submitted 1 February, 2021;
originally announced February 2021.
-
Fractional derivative of composite functions: exact results and physical applications
Authors:
Gavriil Shchedrin,
Nathanael C. Smith,
Anastasia Gladkina,
Lincoln D. Carr
Abstract:
We examine the fractional derivative of composite functions and present a generalization of the product and chain rules for the Caputo fractional derivative. These results are especially important for physical and biological systems that exhibit multiple spatial and temporal scales, such as porous materials and clusters of neurons, in which transport phenomena are governed by a fractional derivati…
▽ More
We examine the fractional derivative of composite functions and present a generalization of the product and chain rules for the Caputo fractional derivative. These results are especially important for physical and biological systems that exhibit multiple spatial and temporal scales, such as porous materials and clusters of neurons, in which transport phenomena are governed by a fractional derivative of slowly varying parameters given in terms of elementary functions. Both the product and chain rules of the Caputo fractional derivative are obtained from the expansion of the fractional derivative in terms of an infinite series of integer order derivatives. The crucial step in the practical implementation of the fractional product rule relies on the exact evaluation of the repeated integral of the generalized hypergeometric function with a power-law argument. By applying the generalized Euler's integral transform, we are able to represent the repeated integral in terms of a single hypergeometric function of a higher order. We demonstrate the obtained results by the exact evaluation of the Caputo fractional derivative of hyperbolic tangent which describes dark soliton propagation in the non-linear media. We conclude that in the most general case both fractional chain and product rules result in an infinite series of the generalized hypergeometric functions.
△ Less
Submitted 8 January, 2019; v1 submitted 7 March, 2018;
originally announced March 2018.
-
Exact results for a fractional derivative of elementary functions
Authors:
Gavriil Shchedrin,
Nathanael C. Smith,
Anastasia Gladkina,
Lincoln D. Carr
Abstract:
We present exact analytical results for the Caputo fractional derivative of a wide class of elementary functions, including trigonometric and inverse trigonometric, hyperbolic and inverse hyperbolic, Gaussian, quartic Gaussian, and Lorentzian functions. These results are especially important for multi-scale physical systems, such as porous materials, disordered media, and turbulent fluids, in whic…
▽ More
We present exact analytical results for the Caputo fractional derivative of a wide class of elementary functions, including trigonometric and inverse trigonometric, hyperbolic and inverse hyperbolic, Gaussian, quartic Gaussian, and Lorentzian functions. These results are especially important for multi-scale physical systems, such as porous materials, disordered media, and turbulent fluids, in which transport is described by fractional partial differential equations. The exact results for the Caputo fractional derivative are obtained from a single generalized Euler's integral transform of the generalized hyper-geometric function with a power-law argument. We present a proof of the generalized Euler's integral transform and directly apply it to the exact evaluation of the Caputo fractional derivative of a broad spectrum of functions, provided that these functions can be expressed in terms of a generalized hyper-geometric function with a power-law argument. We determine that the Caputo fractional derivative of elementary functions is given by the generalized hyper-geometric function. Moreover, we show that in the most general case the final result cannot be reduced to elementary functions, in contrast to both the Liouville-Caputo and Fourier fractional derivatives. However, we establish that in the infinite limit of the argument of elementary functions, all three definitions of a fractional derivative - the Caputo, Liouville-Caputo, and Fourier- converge to the same result given by the elementary functions. Finally, we prove the equivalence between Liouville-Caputo and Fourier fractional derivatives.
△ Less
Submitted 19 February, 2018; v1 submitted 19 November, 2017;
originally announced November 2017.
