XJTLU team develops world’s first independently cross-checked database for RNA modification

In life sciences research, RNA modifications function like “molecular switches” in gene expression, controlling how and when cells produce proteins.

One of the most common RNA modifications is m⁶A（N⁶-methyladenosine), which affects RNA stability and the efficiency of protein synthesis. It has been closely linked to cellular function and even disease development.

For a long time, this field has faced a major challenge: inconsistent results across laboratories.

For the same gene and nucleotide site, one paper may report a modification while another shows none. These inconsistencies make it difficult for researchers to determine which modifications are genuine and reproducible.

Now, this confusion may finally be resolved.

Figure 1 m6AConquer: A data-sharing platform enabling consistent quantification and orthogonal validation across ten m⁶A modification detection technologies

A unified standard

Recently, Dr Zhen Wei and his team from the Department of Biological Sciences and Bioinformatics at Xi’an Jiaotong-Liverpool University, together with researchers from the University of Liverpool, published their latest findings in the international journal Nucleic Acids Research. They introduced m6AConquer, the world’s first database that integrates multiple m⁶A detection techniques and verifies results using orthogonal, independent methods to ensure reproducibility.

“In the past, researchers each had their own experimental system. Now we finally have a shared coordinate system that aligns all the data,” says Dr Wei.

Simply put, m6AConquer consolidates m⁶A sequencing data from 10 experimental technologies and hundreds of public datasets. It employs an algorithm for orthogonal validation, filtering for modification sites that are reproducibly detected across methods with distinct biochemical principles.

This means researchers now have access to a reliable, unified, and verifiable benchmark dataset.

Figure 2 uses orthogonally validated m⁶A sites as ground truth to accurately evaluate the performance of various m⁶A detection methods.

A tool for authenticating RNA modifications

m6AConquer can be seen as a “detector” for verifying RNA modifications.

It helps scientists distinguish between reproducible biological signals and random experimental noise.

The research team adopted the statistical framework used in the internationally recognised ENCODE project. This includes the IDR model, which measures whether different techniques consistently detect the same RNA modification sites, ensuring reliable results. Ultimately, they identified more than 135,000 high-confidence m⁶A modification sites (IDR < 0.05), which demonstrated statistically significant reproducibility across orthogonal methods.

“This is the first time a rigorous ground truth dataset has been established in this field,” says Dr Wei. “It allows researchers to evaluate the performance of various m⁶A detection methods using a shared, validated reference. As a result, there is no longer a need to adjust for methodological differences. This greatly enhances consistency and reproducibility in research and provides a reliable data standard for future algorithm development and disease biology research.”

Figure 3: The Multi-Omics Data-Sharing Framework Developed by m6AConquer

From usable to user-friendly

Beyond ensuring data reliability, Dr Wei’s team also focused on usability – making the data more accessible to scientists.

m6AConquer has established a standardised data-sharing framework that restructures previously scattered and complex multi-omics raw data, capturing multiple layers of information derived from the high-throughput m6A profiling experiments, into “analysis-ready” matrix formats.

This is akin to integrating dictionaries from different countries into one shared language. Researchers no longer need to spend extensive time cleaning and aligning data. They can directly begin comparisons, modelling and validation, significantly improving research efficiency and reproducibility.

A new window into genetic variation and disease mechanisms

Even more promising is that m6AConquer not only integrates RNA modification data but also links RNA modifications to genetic variation, gene expression, and disease risk.

The research team identified over 6,000 genetic variants significantly associated with the modification levels of high-confidence m⁶A sites – also known as m⁶A QTLs. These findings reveal how genetic differences may influence RNA modifications and, in turn, alter gene regulation.

Some of these variants are located in risk regions associated with complex diseases such as psychiatric disorders and depression, offering new insights into their molecular mechanisms.

“This means we’re not only observing the ‘results’ of RNA modifications, but beginning to understand their underlying ‘causes’,” explains Dr Wei. “Some genetic mutations may affect disease risk precisely by altering RNA modification levels.”

Dr Wei adds that m6AConquer is not only a data resource but also a bridge connecting genetic variation, RNA modifications, and disease development. It also lays a solid foundation for future applications such as artificial intelligence modelling, biomarker identification for disease diagnosis, and drug target discovery using high-confidence data.

“We hope this open-access resource will serve as a key step in advancing epitranscriptomics research, the study of chemical modifications on RNA that regulate its function, from data integration to mechanistic understanding,” he says.

The m6AConquer database and all associated analytical tools are freely available to researchers worldwide at: rnamd.org/m6aconquer

Translation:Luyao Wang
Editor:Patricia Pieterse