干细胞,疾病模型 z-bio 2021-04-14 240

　　"No two leaves in the world are the same."

　　Stem cell biologists always suspect that the stem cells cloned into the two genetic materials are exactly the same. Recently, more than 6000 photos of fluorescently labeled induced pluripotent stem cells (iPS) in Seattle have further confirmed the amazing changes in the field of biology, which revealed a lot of basic information about cell biology.

　　On April 5, 2017, the Allen Institute for Cell Science, a subsidiary of The Allen Institute, announced the launch of the "Allen Cell Explorer". This is a unique portal tool for observing human cells for the first time. Dynamic digitization window. The website integrates large-scale 3D imaging data and applies artificial intelligence, machine deep learning and CRISPR gene editing. The system creates visual cellular tissue prediction models and a set of other powerful tools. In addition, the portal and sharing platform allow researchers to predict changes in cell layout, which can herald cancer and other diseases. Dr. Rick Horwitz, Secretary-General of the Allen Institute of Cell Science, said that by revealing unexpected data results at the cellular structure level, this unprecedented tool for stem cell research, cancer research and drug development, he Said that this may accelerate progress. He said: "Cells are extremely complex. They contain thousands of interacting components that work together to drive and regulate the structure and behavior of cells. Launching the AllenCell Explorer website and bringing our grand vision to the world. We are very happy to work with Our researchers share it with cellular data, incredible images, predictive models, etc."

　　The project started about a year ago. Dr. Horwitz and his research team reprogrammed adult skin cells into an undifferentiated embryonic state and used CRISPR-Cas9 technology to insert fluorescent protein "tags." "For genes that clarify the structure of cells. These genes contain genes that code for actin filaments in cells. They are a protein that helps cells move and maintain their shape. Researchers soon discovered that all genetically cloned cells, even Cells from the same parental generation have very different intracellular components, such as the position and shape of mitochondria and actin fibers; IntegratedCellModel is a unique component of AllenCellExplorer, and is the first to apply deep learning technology to predict the intracellular tissue structure of human stem cells Model Creation Model To this end, researchers have "trained" thousands of high-quality images of human stem cells to understand how the composition of stem cells is organized. Specifically, computer scientists use deep learning programs to analyze thousands of images Image and discover the relationship between the positions of structures within the cell. Then, use this information to predict possible positions in the structure, for example when repositioning the nucleus programmatically. The plan aims to compare the predicted results with the actual cells Perform "deep learning".

　　In the next few months, researchers at the Allen Institute will update stem cell images at all stages of cell division. This also means that these cells gradually become different cell types (such as heart cells and kidney cells). Dr. Horwitz believes that capturing cell characteristics at different times is essential to determine its basic development or growth process.

　　Dr. Horwitz continued: "This is the first time researchers have used "deep learning" to try to understand the actual organization of cells. Currently, we mainly rely on textbook schematics. Art scientists. The interpretation of cell data is relatively small. I think that simple charts will eventually be replaced by many unit data-driven models.

　　Allan Jones, Director and CEO of the Allen Institute, said: AllenCell Explorer’s view of human cells plays an extraordinary role. These powerful tools from Allen Institute of Cell Science are some free resources. We always follow the philosophy of creating and sharing powerful open scientific tools and promoting scientific innovation worldwide.