Detecting and analyzing topics of massive COVID-19 related tweets for various countries.

With the flare-up of the COVID-19 infection since 2020, COVID-19 has been one of the hottest topics on Twitter. Topic modeling is one of the most popular analyses, which extracts the topics from the text. This paper proposes a method to extract the most-discussed topics for 32 countries of the world. In this regard, more than five million related tweets have been studied, and a method based on content analysis is proposed to identify the exact location of each tweet. Then, by using the statistical algorithm of Latent Dirichlet Allocation, the main topics of the tweets are identified. By leveraging sentiment analysis, the topics are afterward divided into positive and negative groups, and their trends in a quarterly period are investigated for the countries under study. The outcome of the analysis of time trends shows that for most countries, the trend of negative topics is highly correlated with the number of confirmed cases of COVID-19.

Deep-learning microscopy image reconstruction with quality control reveals second-scale rearrangements in RNA polymerase II clusters.

Fluorescence microscopy, a central tool of biological research, is subject to inherent trade-offs in experiment design. For instance, image acquisition speed can only be increased in exchange for a lowered signal quality, or for an increased rate of photo-damage to the specimen. Computational denoising can recover some loss of signal, extending the trade-off margin for high-speed imaging. Recently proposed denoising on the basis of neural networks shows exceptional performance but raises concerns of errors typical of neural networks. Here, we present a work-flow that supports an empirically optimized reduction of exposure times, as well as per-image quality control to exclude images with reconstruction errors. We implement this work-flow on the basis of the denoising tool Noise2Void and assess the molecular state and 3D shape of RNA polymerase II (Pol II) clusters in live zebrafish embryos. Image acquisition speed could be tripled, achieving 2-s time resolution and 350-nm lateral image resolution. The obtained data reveal stereotyped events of approximately 10 s duration: initially, the molecular mark for recruited Pol II increases, then the mark for active Pol II increases, and finally Pol II clusters take on a stretched and unfolded shape. An independent analysis based on fixed sample images reproduces this sequence of events, and suggests that they are related to the transient association of genes with Pol II clusters. Our work-flow consists of procedures that can be implemented on commercial fluorescence microscopes without any hardware or software modification, and should, therefore, be transferable to many other applications.

Applying Machine Learning for Integration of Multi-Modal Genomics Data and Imaging Data to Quantify Heterogeneity in Tumour Tissues.

With rapid advances in experimental instruments and protocols, imaging and sequencing data are being generated at an unprecedented rate contributing significantly to the current and coming big biomedical data. Meanwhile, unprecedented advances in computational infrastructure and analysis algorithms are realizing image-based digital diagnosis not only in radiology and cardiology but also oncology and other diseases. Machine learning methods, especially deep learning techniques, are already and broadly implemented in diverse technological and industrial sectors, but their applications in healthcare are just starting. Uniquely in biomedical research, a vast potential exists to integrate genomics data with histopathological imaging data. The integration has the potential to extend the pathologist's limits and boundaries, which may create breakthroughs in diagnosis, treatment, and monitoring at molecular and tissue levels. Moreover, the applications of genomics data are realizing the potential for personalized medicine, making diagnosis, treatment, monitoring, and prognosis more accurate. In this chapter, we discuss machine learning methods readily available for digital pathology applications, new prospects of integrating spatial genomics data on tissues with tissue morphology, and frontier approaches to combining genomics data with pathological imaging data. We present perspectives on how artificial intelligence can be synergized with molecular genomics and imaging to make breakthroughs in biomedical and translational research for computer-aided applications.