A research team led by the A*STAR Genome Institute of Singapore (A*STAR GIS) has recently made significant strides in the field of genetic research by developing a method to accurately and efficiently read DNA that contains non-standard bases, a process that was once considered too complex for conventional DNA sequencers. Their work, published in Nature Communications, marries advanced nanopore sequencing technology with artificial intelligence (AI) to decode these extra "letters" swiftly and with high accuracy.

The Challenge: Decoding DNA, Nature's Hidden Language

DNA, the fundamental blueprint of life, is primarily composed of four standard bases, denoted as A, T, C, and G. Scientists have long envisioned the possibility of expanding this genetic alphabet by introducing new characters, referred to as non-canonical bases (NCBs). These NCBs can appear naturally in certain viruses or may be synthetically produced in laboratories. Their inclusion has the potential to revolutionize the design of novel molecules, materials, and biological systems.

Despite their potential, traditional DNA sequencing machines were designed to detect only the four standard bases, which has hindered scientists' capacity to fully exploit these non-standard bases for advancements in precision medicine, synthetic biology, and the creation of innovative materials. Dr. Mauricio Lisboa Perez, a scientist at A*STAR GIS and lead author of the study, elucidates, "Our ability to quickly read a piece of text relies largely on how familiar we are with the vocabulary used. Similarly, for an AI model to 'speed-read' DNA, it must have seen enough examples of every base." However, due to the rarity and complexity of NCBs, generating sufficient data for AI models constituted a significant hurdle.

The Solution: Using AI to Translate DNA with Non-Standard Bases

To tackle this challenge, the researchers constructed an extensive library of artificial DNA featuring both the standard and non-standard bases in a variety of combinations. They employed nanopore sequencing to meticulously capture the unique electrical signals generated as each base traversed microscopic pores. Given the noisy and often incomplete nature of the data collected, the team devised an AI-centric approach that permitted iterative learning, refining its predictive accuracy over time.

Moreover, they augmented the AI model’s educational framework by innovatively rearranging the existing signal data to represent a greater array of possible base combinations. This adaptive strategy empowered the AI to discern and accurately recognize the distinct patterns associated with new DNA "letters," thus enabling the sequencer to read these compounds directly.

Table 1: Comparative Overview of Conventional vs. AI-Enhanced DNA Sequencing

Aspect Conventional Sequencing AI-Enhanced Sequencing
Recognition of Non-Standard Bases Limited, only recognizes A, T, C, G Accurately detects various NCBs
Data Processing Speed Slower, reliant on manual analysis Fast, automated with AI
Scalability Challenging due to training data limitations Easily scalable with iterative learning

Transformative Potential of the Method

The breakthrough achieved by the A*STAR GIS team could catalyze innovation across multiple sectors, primarily:

  • Health Care and Therapeutics: The ability to precisely read and analyze non-standard bases alleviates a critical barrier in the development of DNA and RNA-based treatments, laying the groundwork for pioneering drugs and diagnostics.
  • Advanced Materials and Biotechnology: NCBs could serve as foundational elements for creating nanostructures and nanorobots, fostering advancements in medical applications, manufacturing methodologies, and sustainable chemical production solutions.
  • Data and Information Storage: By employing expanded DNA alphabets for information encoding, the process may become more cost-effective and energy-efficient, thereby potentially mitigating the environmental impact of data centers.

In a statement, Dr. Niranjan Nagarajan, Associate Director of AI & Compute at A*STAR GIS and senior author of the study, remarked, "The ability to accurately identify these new bases at scale provides us with a much richer vocabulary for composing and comprehending biological information. It's akin to learning to recognize new letters, which in turn, enables us to grasp significantly more words and meanings in the language of life."

Table 2: Future Research Directions

Research Focus Description
Discovery of More NCBs Identify additional non-standard bases in various viruses for broader applicability.
AI Model Enhancement Further refine the AI model's capabilities to detect more complex NCBs effectively.
Real-World Applications Investigate practical applications in therapeutics, sustainable production, and nanotechnology.

In conclusion, the innovative DNA sequencing method developed by the A*STAR GIS team stands poised to transform scientific research and application within the realms of health, technology, and beyond. Dr. Wan Yue, Executive Director at A*STAR GIS, excitedly states, "Working with an expanded DNA alphabet will create numerous opportunities for scientists to develop new therapeutics, genetically redesigned organisms for environmental chemical production, and programmable materials for nanostructures and nanorobots. These innovations will drive scientific discovery, generate economic value, and ultimately enhance quality of life."

More information: Mauricio Perez et al, Direct high-throughput deconvolution of non-canonical bases via nanopore sequencing and bootstrapped learning, Nature Communications (2025).

Journal information: This document is subject to copyright. Apart from any fair dealing for the purpose of private study or research, no part may be reproduced without written permission.

For further reading, refer to the citation: Decoding new DNA 'letters' to advance medicine and biotechnology (2025, November 13) retrieved from Phys.org.