Quantum computing for exploring personalized cancer therapy

The latest issue of Nature Biotechnology reported that the German National Cancer Center (DKFZ) is using quantum computing to explore personalized cancer treatment [1].

This summer, the first commercial quantum computer IBM Q system one (27 qubits) in Europe was unveiled in eningen, IBM's headquarters in Germany. It is jointly operated by IBM and Fraunhofer Association, a German multidisciplinary application research organization.

The Fraunhofer society is providing quantum computers to researchers who want to test ideas for practical applications of quantum computers, including in the field of life sciences. Researchers at the German National Cancer Center (DKFZ) will take the lead in testing the potential of this device.

IBM Q system one in einingen, Germany

The most promising application in biomedicine is computational chemistry, and researchers have been using quantum methods for a long time. But the Fraunhofer society hopes to stimulate the interest of a wider group of life scientists, such as cancer researchers. "This is an unknown field," said Dr. Niels halama, head of the translational immunotherapy Department of the German Cancer Research Center (DKFZ) and senior physician of the German National Cancer Disease Center. Halama is working with a team of physicists and computer scientists to plan to develop and test algorithms that may help stratify cancer patients and select small subgroups for specific treatments from heterogeneous data sets.

The medical records of cancer patients can usually contain up to 100 TB of personal data, usually very heterogeneous data, including blood and tumor values, personal indicators, sequencing and treatment data. So far, due to the lack of appropriate processing mechanism, it is impossible to make effective use of this rich information. Therefore, for many cancers, the possibility of using personalized treatment methods promised in the future is still purely theoretical, and patients are still receiving standard treatment.

However, halama said that classical computing does not have enough ability to find very small groups in large and complex data sets generated by oncology, for example. The time required to complete such tasks may last for several weeks - too long to be used in a clinical environment and too expensive. In addition, due to the basic limitation of chip miniaturization, the steady improvement of classical computer performance is slowing down. Although quantum computing is still in its infancy, it may provide a solution to this problem.

"We want to explore how to systematically process and use these heterogeneous data with the help of quantum computers in order to identify new and more targeted options for patients with poor immunotherapy. Finally, we need to know who can benefit from which treatment and how to benefit," halama said.

Sequencing the tumor genome will produce a large amount of data, which may be evaluated by quantum computers in the future.
 

Although there are limited ways to obtain quantum hardware, scientists exploring potential applications have been testing their quantum algorithms on simulators, which are classic high-performance computers programmed to simulate quantum processing.

At present, many people are turning to quantum computing to see if it can help solve atomic or molecular scale problems, such as predicting the size of protein folding, solving enzyme chemistry problems, such as how nitrogenase catalyzes nitrogen fixation, identifying transcription factor binding sites in DNA molecules and performing DNA de novo assembly.

According to halama, there is a huge difference between working on a quantum simulator and working on a real quantum computer (such as IBM Q system one in Germany). Only through the latter can you see how stable things are at a certain degree of complexity, where hidden dangers are and what is possible.

Professor Raoul Klingner, director of the research department of Fraunhofer Association, said: "the application of quantum computing in complex and important fields such as personalized cancer treatment highlights the potential of quantum computing for medicine and many other industries."

The most promising application of quantum computing is in computational chemistry. Pharmaceutical chemists now usually use quantum mechanics to describe interested molecules and their reactions. Quantum computer can simulate molecular dynamics more accurately than classical computer in theory.

Clemens utschig utschig, chief technology officer of Boehringer Ingelheim, headquartered in Mainz, Germany, said that most major pharmaceutical companies are studying the possibility of quantum computing in drug design. In January this year, Boehringer Ingelheim announced a partnership with Google to develop quantum algorithms for this method.

Christopher tautermann, head of computational chemistry at the company, said: "I think this approach will bear fruit in the next few years. We are betting on the future."

It is unclear whether and how quantum computing can play a role in analyzing the large and diverse data sets generated by genomics researchers and neuroscientists. Quantum computers are currently not allowed to input large data sets.

In order to study whether quantum computing can quickly and reliably identify small subgroups of patients who may respond to the same therapy in large and complex data sets (from genome to histology), hamala is trying two different methods.

One involves a machine learning algorithm for quantum processing, which may require a smaller training data set than traditional computing. This method brings hope for human tumor data in the cancer genome map. The other involves designing a new type of algorithm based on different mathematics - topological algebra well handled by quantum computing - to filter the data and find the interesting parts hidden in it.

Halama stressed: "there is no guarantee that the quantum system will provide the solution we want, but there are signs that it is worth following up."

For halama, there are three important criteria for using quantum computers: data protection, speed and flexibility. He said that scientists are still studying test data, but when real patient data is used in the future, according to the agreement between Fraunhofer and IBM, all research projects and user data remain in Germany, and IBM Q system one operates in accordance with Germany's strict data protection law.

Computing speed is another key criterion, because for cancer patients, every day is important and needs to make decisions quickly. The computing speed can make quantum computing better than traditional computing in the future. Since quantum processors can process data in parallel rather than sequentially, they have the potential to analyze large amounts of data in a small part of the time required by ordinary computers.

Halama believes that the flexible monthly ticket model provided by Fraunhofer capability network quantum computing to its partners is also an important factor. "As an academic institution, it allows us to flexibly use the system when needed without investing a huge amount of money for a long time." in addition, Fraunhofer is our important scientific partner, and our cooperation enables us to bridge the gap between theoretical research and applied research for the benefit of patients, "he said.

Charlotte Deane, a computational biologist at the University of Oxford in the UK, said that quantum computing could not solve all the problems that life scientists wanted to solve. "It will speed up a limited number of tasks for us, and we now need to correctly identify these tasks."

She predicted that in a decade or so, "quantum computing will become a useful tool for people like me."

Link:
[1] https://www.nature.com/articles/s41587-021-01116-x
[2] https://www.fraunhofer.de/en/press/research-news/2021/august-2021/personalizing-cancer-treatment-with-quantum-computing.html