According to the World Health Organisation (WHO), cancer is the second leading cause of death worldwide. In 2018 alone, it took the lives of 9.6 million people. 70% of those deaths took place in Africa, Asia and Central and South America.
On top of the devastating effects that the disease have on patients and those around them, the economic impact of cancer is significant and is increasing. Its total annual financial cost in 2010, again according to WHO figures, was estimated at approximately US$ 1.16 trillion.
However lugubrious these facts might be, they are not deterring Ark, a medtech startup funded last year in Singapore, from working “to create a world free from the pain and suffering caused by late-stage cancer”.
If not identified early, cancer patients are diagnosed at late stages when curative treatment may no longer be an option.
By focusing on early cancer detection through the innovation of its own novel analytical methods, Ark is applying big data analytics with the aim to eliminate late-stage cancer-related deaths within the next 30 years.
In this interview with CIO Asia, Dr Zou Ruiyang, co-founder and Chief Technology Officer (CTO) of Ark, spoke to us about the startup’s leading technology innovation and its diagnostic test development.
What is Ark and what is it working on?
Ark was born in early 2018 through the merger of MiRXES, a privately-held life sciences company, and Venturecraft, Southeast Asia’s leading venture capital firm focused on medtech, biotech and digital health.
Last November it raised US$40 million in a record-breaking Series A investment as it targets global expansion and scaling up multi-centre clinical studies focused on early cancer detection.
Dr Ruiyang, who before Ark was Head of Bioinformatics at Singapore’s Agency for Science, Technology and Research (A*STAR) MicroRNA Signature Identification Centre, is primarily credited for Ark’s core microRNA technology and know-how. He oversees all of Ark’s global research and diagnostic test development activities.
By developing sensitive, non-invasive and cost-effective early cancer detection test kits, Ark is harnessing the breakthrough miRNA technology. With a simple blood test, early-stage cancer can be detected even before clinical symptoms appear.
Dr Ruiyang explained us that the startup’s non-invasive blood tests leverage on a patented, best-in-class, micro ribonucleic acid (miRNA) quantification technology developed by their Research and Development (R&D) arm, MiRXES Lab.
“Traditional cancer biomarker tests quantify the amount of certain cancer-associated proteins in blood. However, these traditional biomarkers are not very sensitive, being unable to pick up all instances of cancer, especially in the early stage, and are also not very specific, giving a large proportion of false positive results,” said Dr Ruiyang.
He continued: “Ark’s technology allows us to sensitively and reliably detect miRNA molecules that are novel type of cancer marker which can be used for highly sensitive and specific cancer detection. miRNA biomarkers can be used to detect early cancers more sensitively than other newer blood-based biomarkers that are broadly detected by sequencing technologies.”
What are the challenges in integrating big data for effective patient solutions?
The key to development of any effective patient solution, said Dr Ruiyang, is to find the right parameters to monitor and to build a robust model for prediction.
This is especially true for cancer as the development of the disease is still not fully understood and may vary from person to person. In addition, miRNAs are a relatively recent discovery compared to other types of cancer biomarkers, so having a sufficiently large clinical database that is carefully curated from independent sources is crucial to improving the accuracy of our cancer detection algorithms.
“Clinical data is not as easily obtained as in other big data industries due to the limitations of study design,” he explained CIO Asia. “There are also many pre-analytical and analytical factors that must either be standardised or considered. The challenge is in obtaining quality data, integrating heterogeneous data, and integrating it with advanced analytics for the purpose of improving healthcare outcomes.”
How is Ark innovating novel analytical tools to augment the existing limited data pool and improve the accuracy of its results while reducing false discoveries?
Biomarker analysis and prediction model building rely on machine learning methods that have been built for these purposes.
In Dr Ruyang’s words, Ark has found that the key lies not in developing novel analytical models but in data cleaning and validation, and that’s why they are focusing on developing new tools for these two tasks.
“To improve the data quality, novel normalisation methods and matching methods were utilised to reduce the noise and biases of the data set for supervised and unsupervised learning,” he said. “We also developed novel randomisation and cross-validation methods to ensure the robustness and accuracy of the biomarker discovery and prediction algorithms.”
“Due to the complexity of biological processes, our research and study designs are always guided by our current knowledge of biology,” the CTO added.
How is Ark combining big data analytics and the breakthrough miRNA technology to extract information and improve its product for early cancer detection?
According to Dr Ruiyang, acquiring quality data will always be key for any big data project and particularly so for the scale of diagnostic research they are conducting.
"We believe that combining big data analytics and the understanding of biology is critical to extracting information and improving the product for medical purposes," he explained.
Ark’s products are clinically validated and are based on the largest curated clinical miRNA dataset compiled to date. The enterprise collaborates extensively with top medical research institutions globally and is spearheading Asia’s largest miRNA clinical studies targeting more than 50,000 participants.
"Our clinical miRNA database spans multiple human diseases, including most high incidence cancer types, metabolic diseases, as well as cardiovascular diseases," he told us.
"Our curated database of quality miRNA profiles is key in enabling a global understanding of circulating miRNAs and their role in human health. Data collection will not stop with product launches, however, as we will continue to gather more data and information to improve our products and to gain biological insights into human diseases," concluded Dr Ruiyang.