ARK • Disrupt

In the fight against cancer, single-cell genomics and spatial imaging are game-changing tools that enable researchers to delve into the details of tumor biology. By examining individual cells and their spatial relationships within tumors, scientists are surfacing critical insights that could revolutionize cancer treatment, particularly for B-cell acute lymphoblastic leukemia (B-ALL) and breast cancer.

Addressing B-ALL, researchers from St. Jude Children’s Research Hospital and China’s Hefei Comprehensive National Science Center recently activated single-cell genomics to map the life cycle of B cells. They discovered that in some developmental stages, particularly the pre-pro-B stage, B-ALL is resistant to asparaginase, a chemotherapy drug. That breakthrough led them to combine asparaginase with venetoclax, a drug that interferes with a protein that helps cancer cells survive, a combination that improved treatment efficacy significantly and now offers hope that patients with B-ALL will overcome the all-too-often resistance to drugs.¹

Similarly, researchers at Fynn Biotechnologies applied 10X Genomics’ Xenium In Situ single-cell spatial imaging to study the tumor microenvironment (TME) in various breast cancers. The technology enabled them to map the organization of and interaction among different cells within the TME, highlighting that the spatial arrangement of immune cells relative to cancer cells plays a crucial role in the effectiveness of immunotherapy. In patients who responded well to treatment, for example, immune cells interacted closely with cancer cells, while the cellular organization associated with non-responders was different and much less effective.²

These findings underscore the profound impact of single-cell genomics and spatial methods in advancing our understanding of cancer. By providing detailed, cell-by-cell views of tumor biology, new technologies are enabling researchers to identify new biomarkers, unravel the complexities of resistance to treatment, and develop more personalized and effective cancer therapies. The ability to integrate cellular with spatial data is a leap forward in precision medicine, bringing patients closer to tailored treatments that can improve outcomes in even the most challenging cancer cases.

The integration of Large Language Models (LLMs) with digital wallets like Apple Pay, WeChat Pay, and Shop Pay could create “smart wallets”—powerful e-commerce aggregators that potentially would transform e-commerce.

Integrating seamlessly with LLMs, digital wallets should be able to understand the intent behind consumer purchases, enabling them to offer consumers valuable insights with advanced search functionality. As a result, digital wallets could become central ecommerce aggregators facilitating lead generation and ultimately streamlining the purchase process.

According to a recent survey conducted by PYMNTS,³ 80% of consumers use stored credit/debit cards and other credentials during e-commerce purchases, underscoring the need for seamless checkout options and efficient payment processes. By offering a seamless, unified solution for both lead generation and payments, digital wallets could increase their share of global e-commerce purchase volume from 50% today to 72% by 2030, as shown below.

Note: The dotted purple line refers to ARK estimates for digital wallets as a payment method of global e-commerce purchase value. The 2018-23 figures (solid purple line) are historical figures documented in the Worldpay Global Payments Reports, and the 2024-30 figures (dotted purple line) are ARK’s E-com S-curve adoption estimates. Source: ARK Investment Management LLC, 2024, based on data from WorldPay as of August 30, 2024.⁴ Forecasts are inherently limited and cannot be relied upon. For informational purposes only and should not be considered investment advice or a recommendation to buy, sell, or hold any particular security or cryptocurrency.

Given the conversion of digital wallets into LLM-enabled smart wallets, consumers could engage with their digital wallets much earlier during the e-commerce purchase cycle than is the case today. Today, especially in the US, consumers interact with their digital wallets later in the process, typically at checkout. Now, leveraging AI agents to compare selection, prices, delivery fees, and delivery times across marketplaces, consumers might delegate most of their e-commerce shopping experiences— from search to checkout—to their smart wallets by 2030.

On August 24, the National Anti-Fraud Office (ONAF) in France arrested⁵ Telegram Messenger CEO Pavel Durov at Le Bourget Airport, alleging⁶ that Telegram was complicit in illegal activities including fraud, money laundering, and the distribution of material on child sexual abuse. Although the authorities released him under judicial supervision for $5.56 million euros⁷ in bail, Durov must remain in France.

While Durov’s arrest might have been a necessary step toward holding platform operators accountable for illegal activities, the case also raises concerns about overreach by French authorities. Telegram Messenger has argued that the platform should not be held responsible for the actions of its users, warning that Durov’s arrest could set a dangerous precedent for freedom of speech, encryption technology, and innovation. The case is likely to set an important precedent around a platform’s liability for user actions, with implications for innovation, in the European Union.

Terray Therapeutics is making significant strides in the use of artificial intelligence (AI) to revolutionize drug discovery. Recently, the company introduced COATI-LDM, a new method that leverages latent diffusion models to design and optimize small molecules critical in the drug-discovery process. Traditional drug discovery is slow and expensive, requiring the careful balancing of properties like potency, safety, and bioavailability. Terray's approach addresses those challenges by using AI to navigate complex molecular "blueprints," generating new candidates that meet specific therapeutic needs.⁸

The COATI-LDM model is trained on vast datasets that help it generate molecules with the right properties for targeting specific diseases. By operating on encoded versions of molecules—rather than on their actual structures—the model can explore chemical space to identify the most promising candidates, not only minimizing the need for expensive and time-consuming laboratory experiments but also accelerating the process and reducing costs significantly.

Terray’s AI-driven approach enables faster drug discovery by generating potential drug candidates that are more likely to succeed in clinical trials. Additionally, Terray’s approach allows for better, more precise design, tailoring molecules to meet specific medical needs. Reduced reliance on traditional, trial-and-error laboratory methods also makes drug development more cost-effective, potentially lowering the price of new medications. Finally, Terray Therapeutics' innovation marks a significant advancement in the approach to drug discovery, promising a future where life-saving treatments are developed faster, cheaper, and more accurately.