Chemists at the Massachusetts Institute of Technology have developed a new model that can quickly predict the three-dimensional genome structure using generative artificial intelligence (AI) technology. The new model can predict thousands of structures in just a few minutes, much faster than existing experimental analysis methods. The research results were published in the latest issue of the journal Science Advances. Every cell in the human body contains the same genetic sequence, but each cell only expresses a portion of its genes. This cell specific gene expression pattern ensures differences between different types of cells, and these expression patterns depend in part on the three-dimensional structure of genetic material. The new model is called ChromoGen, which aims to predict the three-dimensional genome structure from the underlying DNA sequence. It consists of two components, the first being a deep learning model that can "read" the genome and analyze the information encoded in DNA sequence and chromatin accessibility data; The second part is a generative AI model that can predict physically accurate chromatin conformation after training. The dataset used to train the model includes over 11 million chromatin conformations, sourced from experiments conducted on 16 human B lymphocyte lineage single cells. The ChromoGen model can effectively capture sequence structure relationships and generate multiple possible structures for each DNA sequence. This is because DNA is a very disordered molecule, and the same DNA sequence can produce many different conformations. This method not only greatly accelerates the prediction process of three-dimensional genome structure, but also provides a powerful tool for in-depth research on how the three-dimensional organization of the genome affects the gene expression patterns and functions of cells. This breakthrough technology is expected to help scientists understand gene regulation mechanisms faster, promoting advances in disease treatment and biotechnology. (New Society)