Hybrid quantum-classical algorithm for analyzing many-body systems: from NMR inference to validating microscopic theories on quantum simulators

Abstract

Publically Releaseable The goal of this program is to develop hybrid approaches that combine methods and techniques of quantum simulations, quantum computing, and classical machine learning to solve fundamental problems in biomedical and physical sciences. We will address the following specific problems: i) quantum assisted NMR inference for metabolomics, ii) analysis of quantum states using snapshots of many-body wavefunctions obtained in quantum simulators, iii) variational quantum-classical hybrid solution of interacting many-body systems of electrons and phonons. NMR is one of the most powerful analytical techniques available to medicine and biology, as it is suited for both in vivo and in vitro studies. On the other hand, NMR experiments face the challenge of interpreting the data. One directly observes only the magnetic spectrum of a biological sample, whereas the ultimate goal is to learn the underlying microscopic Hamiltonian and ultimately identify and quantify chemical compounds. This inference is tractable for small molecules, but quickly becomes problematic when the complexity of the molecules increases or many spectra overlap. We will develop hybrid quantum-classical approaches to the NMR inference problem, which utilize quantum computers to simulate spin Hamiltonians and obtain the spectra, while the inference part is solved on classical computers. We will develop protocols optimized for different types of quantum hardware, address questions of scalability to larger molecules, understand fundamental limitations of the method. A crucial feature of quantum simulators is the possibility of taking snapshots of many-body states, i.e. simultaneous measurement of every particle in the system. We will develop machine learning protocols that compare many-body images generated by specific theoretical models to experimental results from the atomic Fermi gas microscopes. When several alternative models are available, we will use tools of Data Science to determine which of them provide a better description of experimental data. Variational Quantum Eigensolvers provide a new hybrid variational classical-quantum approach for analyzing interacting systems of electrons and spins. We will develop a family of VQE protocols that can be used to study electron-phonon systems. Our approach will be based on combining classical partial polaron transformations, which entangle phonon and electron degrees of freedom, with VQE approach for optimizing the electronic part of the wavefunction.

Document Details

Document Type
DoD Grant Award
Publication Date
Jul 09, 2020
Source ID
W911NF2010163

Entities

People

  • Eugene A. Demler

Organizations

  • Army Contracting Command
  • Harvard University
  • United States Army

Tags

Fields of Study

  • Physics

Readers

  • Distributed Systems and Data Platform Development
  • Quantum spin resonance or Electron Paramagnetic Resonance spectroscopy.

Technology Areas

  • AI & ML
  • AI & ML - Machine Learning Algorithms
  • Biotechnology
  • Microelectronics
  • Quantum Computing
  • Quantum Science - Quantum Dots