Researchers at Carnegie Mellon University’s School of Computer Science have made a big step forward in understanding how the human genome is organized inside a single cell. This understanding is essential for studying how DNA structure affects gene expression and disease processes. The findings were published in the journal “Nature Methods.”m>, Jian Ma, and former Ph.D. students Kyle Xiong and Ruochi Zhang have developed a machine learning method called scGHOST. This method can detect subcompartments, which are a specific type of 3D genome feature in the cell nucleus, and link them to gene expression patterns.
In human cells, chromosomes are not arranged in a linear manner. Instead, they are folded into 3D structures. Researchers are particularly interested in 3D genome subcompartments because they reveal the spatial locations of chromosomes within the nucleus.
The ultimate goal of single-cell biology is to understand the links between these 3D genome structures and gene expression patterns.Cellular structure and function vary widely in different biological situations,” Ma said. “We’re investigating how chromosome organization within the nucleus is related to gene expression.”
New technologies make it possible to study these structures at the single-cell level, but poor data quality can make it difficult to understand them accurately. To address this issue, scGHOST uses graph-based machine learning to improve the data, making it easier to determine and identify the spatial organization of chromosomes. scGHOST is based on the Higashi method developed by Ma’s research group.
By accurately identifying 3D geAn innovative new tool called scGHOST has been developed to aid scientists in analyzing single cells and the complex molecular makeup of tissues, particularly in the brain. This tool, which focuses on identifying single-cell 3D genome subcompartments, is anticipated to provide valuable insights into gene regulation in both health and disease.
