This dataset contains the bright-field live cell images of murine immune-related cell lines (1008 cells) and primary murine peripheral blood mononuclear cells (PBMCs) (346 cells). Each cell has 30 frames that were captured in 29 min with 1 min interval. The cells of murine cell lines were determined as T cells (318 cells), leukemia cells (423 cells), or hematopoietic progenitor cells (267 cells) based on their subsequently measured whole transcriptomes. The PBMCs were determined as CD4+ T cells (152 cells), CD8+ T cells (103 cells), or B cells (91 cells) based on their subsequently measured whole transcriptomes. The determined cell types are indicated in an accompanying txt file.
Iamges were captured using an inverted optical microscope (Eclipse Ti2-E, Nikon) with an 20X objective lens (CFI Plan Apochromat Lambda 20X, Nikon), an LED (TI2-D-LHLED, Nikon), and a CMOS camera (Zyla5.5, Andor).
Jianshi Jin, Taisaku Ogawa, Nozomi Hojo, Kirill Kryukov, Kenji Shimizu, Tomokatsu Ikawa, Tadashi Imanishi, Taku Okazaki, Katsuyuki Shiroguchi (2023) Robotic data acquisition with deep learning enables cell image-based prediction of transcriptomic phenotypes., Proceedings of the National Academy of Sciences of the United States of America, Volume 120, Number 1, pp. e2210283120
Published in 2023 Jan 3 (Electronic publication in Dec. 28, 2022, midnight )
(Abstract) Single-cell whole-transcriptome analysis is the gold standard approach to identifying molecularly defined cell phenotypes. However, this approach cannot be used for dynamics measurements such as live-cell imaging. Here, we developed a multifunctional robot, the automated live imaging and cell picking system (ALPS) and used it to perform single-cell RNA sequencing for microscopically observed cells with multiple imaging modes. Using robotically obtained data that linked cell images and the whole transcriptome, we successfully predicted transcriptome-defined cell phenotypes in a noninvasive manner using cell image-based deep learning. This noninvasive approach opens a window to determine the live-cell whole transcriptome in real time. Moreover, this work, which is based on a data-driven approach, is a proof of concept for determining the transcriptome-defined phenotypes (i.e., not relying on specific genes) of any cell from cell images using a model trained on linked datasets.(MeSH Terms)