QAOA parameter initialization to avoid barren plateaus (“hardware efficient ansatz design”) - ON-166

Preferred Disciplines: PhD in Quantum Information or appropriate field (Theoretical Physics, Computer Science, Statistics or Mathematics)
Company: ProteinQure Inc.
Project Length: 4-6 months (1 unit)
Desired start date: ASAP
Location: Toronto, ON
No. of Positions: 1
Preferences: Language: English

About the Company: 

ProteinQure (Toronto based startup) is a software platform for computational peptide drug discovery. We combine quantum computing, molecular simulations and machine learning to do the structure based design of drugs. A physics-based approach is less data-dependent and enable us to develop therapeutics for complicated disease targets. We are one of the top graduates of the Quantum Machine Learning stream at the Creative Destruction Lab incubator (University of Toronto) and have partnerships with several quantum computing hardware providers (Rigetti, IBM, Xanadu, D-Wave and Fujitsu).

Project Description:

ProteinQure is working on combining cutting edge computational technologies to help with the design of protein based therapeutics.

One of our major goals is to use the results obtained from noisy quantum computations to improve and speed up molecular dynamics simulations on classical hardware. For example, one of our quantum algorithms models protein folding as an optimization problem on a discrete lattice. In our most recent work, we adapted the Quantum Approximate Optimization Algorithm (QAOA) to solve the protein folding problem on universal gate based quantum computers. We currently have access to both, the IBM and Rigetti quantum computers.

The QAOA algorithm has become very popular among researchers due to its applicability and usefulness in the near-term. In the next 3-5 years, researchers expect noisy intermediate scale quantum computers without error correction rather than fault-tolerant devices. The QAOA belongs to the class of variational quantum algorithms which have shown to be robust against certain types of errors. It is expected that quantum advantage (often called quantum supremacy) will be demonstrated with such a variational algorithm in the near future. However, a known problem with this algorithm is that choosing “good” initial parameters becomes more difficult as the quantum computers scale. Current implementations often start with random parameter initialization. 

The research question for this Mitacs proposal is how to best choose the initial parameterization for QAOA in order to avoid barren plateaus when it is being applied to combinatorial optimization problems. With the specific example of protein folding providing a real world case study. This would address and ideally resolve the concerns outlined in the aforementioned paper. Any progress in this area will have big impact on the scalability of variational approaches to near-term quantum computing in general.

Research Objectives:

  • Determine suitable method for choosing initial parametrization of QAOA based on the energy landscape of the chosen problem
  • Benchmark different approaches
  • Test strategy on the problem of lattice protein folding

Methodology:

  • Consider how to translate information from energy landscapes to choice of initial parameters
  • Create algorithm (or strategy) for choosing initial parametrization given a problem
  • Consider specific example in protein folding
  • Use quantum computers and quantum computer simulators to benchmark the approach vs other strategies (such as random initialization). QPU access will be provided by ProteinQure

Expertise and Skills Needed:

    • Familiarity with quantum computing algorithms
    • Programming experience

    For more info or to apply to this applied research position, please

    1. Check your eligibility and find more information about open projects
    2. Interested students need to get the approval from their supervisor and send their CV along with a link to their supervisor’s university webpage by applying through the webform.
    Program: