Researchers have found a complementary way of interpreting changes in non-coding DNA that may contribute to disease. Many genetic changes associated with human disease lie outside protein-coding genes, in nearby regulatory DNA that controls when and how genes are expressed.These changes can alter how transcription factors - the proteins that control when genes turn on - attach to DNA. Predicting the effects of such changes is difficult because most computer tools rely only on DNA sequence patterns and cannot show how a transcription factor physically interacts with DNA. 3D modelling methods Researchers at the Institute of Genetics and Cancer (IGC) tested whether new 3D modelling methods, such as AlphaFold 3, can help explain these subtle effects. By modelling thousands of protein–DNA pairs and comparing them to high-quality experimental data, they found that structural models can often tell which version of a DNA sequence a transcription factor prefers to bind. This approach provides clear mechanistic insight by showing how a single DNA change can strengthen or weaken binding and provides a complementary way to identify and understand non-coding variants that may contribute to human disease. Getting at the effect of genetic variation in the non-coding genome is complicated. If we can generate data to improve 3D modelling methods, they could be a useful new tool to use in conjunction with other methods. Professor Joe Marsh Group Leader, Institute of Genetics and Cancer Read the full paper in Nucleic Acids Research Joe Marsh Research Group Tags 2026 Publication date 12 Jan, 2026
