Inside Short
- The Cleveland Clinic-IBM team demonstrated a hybrid quantum-classical workflow to estimate the electronic structure of the 303-atom Trp-cage protein using the IBM Quantum Heron r2.
- The process combines wavelet-based processing to classify proteins into groups and sample-based quantum diagonalization to solve complex electronic structures.
- The approach goes beyond the Trp-cage and can support pharmaceutical research and molecular simulations using quantum-centric supercomputing.
A joint Cleveland Clinic-IBM team has used quantum computing to demonstrate a hybrid quantum-classical flow to estimate the electronic structure of a protein for the first time, IBM announced. The team modeled the 303-atom miniprotein Trp-cage using a quantum-centric supercomputing workflow with the IBM Quantum Heron r2.
The work represents progress in designing and implementing quantum-centric supercomputing algorithms and workflows that combine quantum and high-performance classical computing.
Accurate calculations of the electronic structure in older computers become more difficult as the system size increases, the company said. Only primitive methods can accurately mimic certain aspects of protein behavior, but highly accurate treatments of complete proteins remain elusive.
Example of a Trp-Cage Miniprotein
Trp-cage facilitates benchmarking methods in computational chemistry. The molecule is relatively compact for a protein, but it has features common to larger biochemistry molecules, such as a hydrophobic core and hydrogen bonding between its parts, which allows it to take more complex forms. The researchers modeled its relaxed and folded states.
“Proving that this method works for Trp-cage is a step towards larger molecules,” said Mario Motta, co-author of the paper, as quoted by IBM.
The team originally planned to mimic a few amino acids, Motta said. But when they tested their process, they found that they were able to reach the Trp-cage and got meaningful results.
Dr. Kenneth Merz, who leads the Merz lab at the Cleveland Clinic, said he hopes these methods can support computational work for medical research and related fields as they grow and develop. He envisions a world where scientists use quantum-centric supercomputing workflows to build databases of molecular behavior, then use machine learning algorithms trained on those databases to identify molecules that can behave in the desired way.
Wave Function-Based Embedding Workflow
The process, described in the latest publication on arXiv, is based on a method called wave function-based embedding to fragment Trp-cage into possible pieces called “clusters,” the source reported. At this point, there are as many groups as there are atoms in a molecule, but each group is more complex than a single atom, involving a local region surrounding the atom and being held by it.
Some groups are more complex than others, the company explained. A single atom can be at the end of a protein and is only attached to one or two neighboring atoms, which allows researchers to find the electronic structure of the cluster efficiently using traditional computer methods. One may be closer to the molecular backbone, embedded in a complex network of molecular interactions. These large groups are good problems for quantum computers to solve.
Taken together, the mathematical results of the individual groups lead to a complete solution for the electronic structure of the molecule, which describes where its electrons are and how they interact.
Sample-Based Quantum Diagonalization Algorithm
Merz has been eyeing the development of a quantum computer for several years, according to IBM. Until a few years ago, it was clear that quantum computers could provide new ways to solve complex chemistry problems, but what those ways would look like was still an open question.
Merz says there was a eureka moment when he saw IBM scientists develop an algorithm called sample-based quantum diagonalization (SQD). The algorithm belongs to an emerging group of algorithms built for quantum-centric supercomputing, where classical and quantum devices work together to solve problems using the strengths of both paradigms, as reported by IBM.
Sample-based quantum diagonalization (SQD) addresses one of the fundamental challenges of electronic structure calculations – the number of possible configurations of a molecule’s electrons grows in proportion to the size of the molecule. Algorithmically, a quantum computer models this vast space, showing the basic settings for a classical computer to focus on. A primitive computer uses the information that appears to find a solution.
“We dropped everything. I met a few people in my group over the weekend, and we decided to go all in on SQD,” Merz said.
The team first tested the algorithm on a series of small molecules, starting the series of experiments that led to this Trp-cage simulation. The results so far have been very promising, and the work is already working in competition with the old methods and approaching the accuracy of the most difficult things among them, the source reports.
In fact, a combined wave-based embedding and sample-based quantum diagonalization workflow can go far beyond the Trp-cage, the scientists said. As molecules grow, the task of breaking them apart, reading their most complex groups, and putting them back together becomes more difficult. But solving the electronic structure of complex molecules is a formidable problem for quantum computers. Researchers are exploring what the next step looks like, targeting larger molecules.
The work was made possible by access to high-performance computing facilities at Michigan State University and the Cleveland Clinic, IBM said. Other recent collaborations between IBM and HPC leaders such as RIKEN have also yielded results.
Source – Cleveland Clinic and IBM launch new quantum process for protein modeling.
#Cleveland #Clinic #IBM #Researchers #Simulate #Protein #Structures #Quantum #Computing