Mr Mark Fingerhuth not only earned a cum laude result for his Master’s in Physics but has also been recognised by Forbes and the Massachusetts Institute of Technology (MIT) for his innovative and cutting-edge Canada-based biotechnology startup, ProteinQure, which is advancing computational drug discovery in the pharmaceutical industry.
Fingerhuth’s academic career began in The Netherlands, where he studied for a Bachelor of Science in liberal arts and the sciences, majoring in theoretical physics, at the Maastricht Science Programme, part of Maastricht University.
His institution encouraged students to spend six months with a research group worldwide to work on their undergraduate thesis topic, leading Fingerhuth to discover the work of Professor Francesco Petruccione and Dr Maria Schuld at UKZN. Having lived in Ethiopia for a year and enjoying time on the African continent, Fingerhuth was excited at the prospect of joining UKZN’s Centre for Quantum Technology, citing many of their innovative research publications when he contacted Petruccione, resulting in an invitation to join their research group.
Fingerhuth’s fascination with theoretical physics stems from physicists’ impressive ability to derive the laws of nature using just a blackboard and chalk, making accurate predictions about experiments even outside a laboratory, an especially inspiring feat for fields like quantum mechanics, general relativity, and particle physics, where theories like the Standard Model, developed in the 1970s, continue to predict phenomena and explain experimental results decades later successfully.
Quantum computing was particularly interesting to Fingerhuth because of physicists and technologists’ seemingly miraculous ability to teach raw materials like silicon and metal to perform complex computations. Intrigued by assembly language, quantum computing attracted Fingerhuth as quantum programmers must still operate less abstractly and consider the practical assembly of materials and algorithmic designs.
His master’s thesis concerned the learning curves of quantum kernel machines. Quantum Machine Learning (QML) develops machine-learning algorithms for quantum computers, and Fingerhuth focused on quantum-classical hybrid models, which perform some computations on classical computers and others on quantum processors to analyse how the size of a training dataset affects the accuracy of QML models, known as a “learning curve”. He sought to address whether researchers can predict which QML model will perform best on a given dataset without trial and error. Using learning curves and statistical physics tools, he investigated whether a model’s performance can be theoretically calculated, avoiding the need to train multiple models.
‘This calls back to why I am fascinated with theoretical physics; it’s the ability to analytically calculate things and make predictions about the real world without having to do the trial-and-error experiments in the real world,’ said Fingerhuth.
His results provided some interesting insights, including the direct relation between the Fourier spectrum of a (stationary) quantum model and the number of quantum bits (qubits) used to implement that model, which explained why adding more qubits leads to a more expressive QML model. Matching the Fourier spectrum of the dataset to that of the QML model results in the best performance, a rational result as a QML model would not easily be able to learn a dataset containing a Fourier frequency the model did not possess.
Fingerhuth completed this complex research while working full-time at ProteinQure in Toronto, Canada, resulting in many late nights and working weekends. The stressful juggle of work and study paid off, as his efforts were recognised with excellent academic results and his start-up’s efforts in peptide-based drug delivery have paid dividends.
Chief Research and Development (R&D) Officer at ProteinQure, over the past seven years since its initiation, Fingerhuth has been successful in raising more than $15 million in venture capital funding, has signed multiple deals with AstraZeneca and several other Top20 pharma companies, and has seen ProteinQure’s most promising drug against the deadly triple-negative breast cancer poised to go to Phase I clinical trials in 2025.
Fingerhuth was named a Forbes 30 Under 30 visionary entrepreneur and an MIT Innovator Under 35. He manages a team of four Machine Learning scientists and engineers and oversees ProteinQure’s R&D efforts regarding computational drug discovery. The company’s projects include developing evolutionary algorithms, building peptide diffusion models, new embedding spaces for non-natural amino acids, and more. They recently received a $700k grant from the Canadian Government to investigate how quantum computing can impact computational drug design in the future.
Fingerhuth thanked Petruccione and Schuld for their effort and support during his studies, expressing deep appreciation for the profound impact they had on his academic and personal development and for the lifelong friendships he formed with them. He thanked Ms Nelisiwe Mncube and Mr Mpho Mkhwanazi for their responsiveness, patience, and understanding, and for helping him navigate UKZN’s complex systems to graduate on time.
Words: Christine Cuénod
Photograph: Supplied
