Stanford Center for Biomedical Informatics Research (BMIR), Stanford University, School of Medicine, 1265 Welch Road, Stanford, CA 94305-547, USA.
Computer Engineering, Automatics and Robotics Department, University of Granada, C. Periodista Daniel Saucedo Aranda, s/n, Granada, 18014 Granada, Spain.
Cell Rep Methods. 2023 Jul 19;3(8):100534. doi: 10.1016/j.crmeth.2023.100534. eCollection 2023 Aug 28.
In this work, we propose an approach to generate whole-slide image (WSI) tiles by using deep generative models infused with matched gene expression profiles. First, we train a variational autoencoder (VAE) that learns a latent, lower-dimensional representation of multi-tissue gene expression profiles. Then, we use this representation to infuse generative adversarial networks (GANs) that generate lung and brain cortex tissue tiles, resulting in a new model that we call RNA-GAN. Tiles generated by RNA-GAN were preferred by expert pathologists compared with tiles generated using traditional GANs, and in addition, RNA-GAN needs fewer training epochs to generate high-quality tiles. Finally, RNA-GAN was able to generalize to gene expression profiles outside of the training set, showing imputation capabilities. A web-based quiz is available for users to play a game distinguishing real and synthetic tiles: https://rna-gan.stanford.edu/, and the code for RNA-GAN is available here: https://github.com/gevaertlab/RNA-GAN.
在这项工作中，我们提出了一种使用深度生成模型和匹配的基因表达谱生成全切片图像（WSI）瓦片的方法。首先，我们训练一个变分自编码器（VAE），它学习多组织基因表达谱的潜在、低维表示。然后，我们使用这个表示来注入生成对抗网络（GANs），生成肺和大脑皮层组织瓦片，从而产生我们称之为 RNA-GAN 的新模型。与使用传统 GAN 生成的瓦片相比，由 RNA-GAN 生成的瓦片更受专家病理学家的青睐，此外，RNA-GAN 需要更少的训练周期来生成高质量的瓦片。最后，RNA-GAN 能够推广到训练集之外的基因表达谱，显示出插补能力。一个基于网络的测验可供用户玩一个游戏，区分真实和合成的瓦片：https://rna-gan.stanford.edu/，并且 RNA-GAN 的代码可在此处获得：https://github.com/gevaertlab/RNA-GAN。
