[Paper Review] Tangelo: An Open-source Python Package for End-to-end Chemistry Workflows on Quantum Computers

Valentin Senicourt, James M. Brown+11 more|arXiv (Cornell University)|Jun 24, 2022

Quantum Computing Algorithms and Architecture4 citations

TL;DR

Tangelo is an open-source Python package that enables end-to-end quantum chemistry workflows on quantum computers by providing modular, backend-agnostic tools for designing, compiling, and executing quantum circuits. It supports problem decomposition, variational quantum algorithms like VQE, and integration with multiple quantum backends (e.g., Qiskit, Cirq, Braket), significantly lowering resource requirements and accelerating experimentation for molecular systems including open-shell and excited states.

ABSTRACT

Tangelo [link: https://github.com/goodchemistryco/Tangelo] is an open-source Python software package for the development of end-to-end chemistry workflows on quantum computers, released under Apache 2.0 license. It aims to support the design of successful experiments on quantum hardware, and to facilitate advances in quantum algorithm development. The software enables quick exploration of different approaches by assembling reusable building blocks and algorithms, with the flexibility to let users introduce their own. Tangelo is backend-agnostic and enables switching between various backends (Braket, Qiskit, Qulacs, Azure Quantum, QDK, Cirq...) with minimal changes in the code. The package can be used to explore quantum computing applications such as open-shell systems, excited states, or more industrially-relevant systems by leveraging problem decomposition at scale. This paper outlines the design choices, philosophy, and main features of Tangelo.

Motivation & Objective

• To address the challenge of designing practical, resource-efficient quantum chemistry experiments on near-term quantum hardware.
• To reduce the barrier for researchers to explore and compare quantum algorithms by providing reusable, composable building blocks.
• To enable seamless integration across multiple quantum computing backends (e.g., Qiskit, Cirq, Azure Quantum) with minimal code changes.
• To support advanced workflows involving problem decomposition (e.g., DMET), variational algorithms, and measurement optimization for accurate results.
• To foster community-driven development and interoperability through open-source collaboration and modular software design.

Proposed method

• Uses a modular architecture with reusable components for molecular input, Hamiltonian transformation, and quantum circuit construction.
• Employs second-quantized molecular Hamiltonians expressed via FermionOperator and mapped to qubit operators using fermion-to-qubit mappings (e.g., Jordan-Wigner, Bravyi-Kitaev).
• Supports variational quantum eigensolver (VQE) workflows with parameterized quantum circuits and backend-agnostic execution via a unified interface.
• Enables circuit manipulation through operations like concatenation, inversion, splitting into non-entangled subsystems, and stacking for parallel execution.
• Integrates with third-party tools like PySCF for Hartree-Fock calculations and integral transformation, and OpenFermion for operator handling.
• Provides measurement protocol optimization by identifying only real-coefficient terms in the qubit Hamiltonian to minimize shot count and improve accuracy.

Experimental results

Research questions

• RQ1 How can end-to-end quantum chemistry workflows be designed to minimize qubit and gate count while maintaining accuracy on near-term quantum devices?
• RQ2 What is the most effective way to decompose large molecular systems into smaller, tractable fragments for quantum simulation?
• RQ3 How can quantum circuit design be abstracted and made portable across multiple quantum computing backends with minimal code changes?
• RQ4 What role do pre- and post-processing techniques play in reducing measurement overhead and improving convergence in variational quantum algorithms?
• RQ5 How can reusable, composable software components accelerate the development and comparison of new quantum algorithms for chemistry?

Key findings

• Tangelo enables the construction of end-to-end quantum chemistry workflows with minimal boilerplate, supporting rapid prototyping of quantum algorithms across multiple backends.
• The package successfully supports problem decomposition via DMET, reducing resource requirements for large systems like H10 by isolating fragments for individual quantum simulation.
• Circuit manipulation features such as splitting, stacking, and variational gate control allow for optimized resource usage and parallel execution strategies.
• The backend-agnostic design allows seamless switching between Qiskit, Cirq, Braket, and other frameworks with minimal code changes, enhancing portability and reproducibility.
• Tangelo’s integration with PySCF and OpenFermion enables accurate molecular Hamiltonian generation and transformation, forming a robust foundation for quantum chemistry simulations.
• The tool supports variational quantum algorithms like VQE with parameterized circuits and provides tools to extract and analyze results, including energy computation from measured expectation values.