Although RNA-guided genome editing via the CRISPR-Cas9 system is now widely used in biomedical research, genome-wide target specificities of Cas9 nucleases remain controversial. Here we present Digenome-seq, in vitro Cas9-digested whole-genome sequencing, to profile genome-wide Cas9 off-target effects in human cells. This in vitro digest yields sequence reads with the same 5' ends at cleavage sites that can be computationally identified. We validated off-target sites at which insertions or deletions were induced with frequencies below 0.1%, near the detection limit of targeted deep sequencing. We also showed that Cas9 nucleases can be highly specific, inducing off-target mutations at merely several, rather than thousands of, sites in the entire genome and that Cas9 off-target effects can be avoided by replacing 'promiscuous' single guide RNAs (sgRNAs) with modified sgRNAs. Digenome-seq is a robust, sensitive, unbiased and cost-effective method for profiling genome-wide off-target effects of programmable nucleases including Cas9.

Several CRISPR-Cas9 orthologues have been used for genome editing. Here, we present the smallest Cas9 orthologue characterized to date, derived from Campylobacter jejuni (CjCas9), for efficient genome editing in vivo. After determining protospacer-adjacent motif (PAM) sequences and optimizing single-guide RNA (sgRNA) length, we package the CjCas9 gene, its sgRNA sequence, and a marker gene in an all-in-one adeno-associated virus (AAV) vector and produce the resulting virus at a high titer. CjCas9 is highly specific, cleaving only a limited number of sites in the human or mouse genome. CjCas9, delivered via AAV, induces targeted mutations at high frequencies in mouse muscle cells or retinal pigment epithelium (RPE) cells. Furthermore, CjCas9 targeted to the Vegfa or Hif1a gene in RPE cells reduces the size of laser-induced choroidal neovascularization, suggesting that in vivo genome editing with CjCas9 is a new option for the treatment of age-related macular degeneration.

The main obstacle to weakly supervised semantic image segmentation is the difficulty of obtaining pixel-level information from coarse image-level annotations. Most methods based on image-level annotations use localization maps obtained from the classifier, but these only focus on the small discriminative parts of objects and do not capture precise boundaries. FickleNet explores diverse combinations of locations on feature maps created by generic deep neural networks. It selects hidden units randomly and then uses them to obtain activation scores for image classification. Fick-leNet implicitly learns the coherence of each location in the feature maps, resulting in a localization map which identifies both discriminative and other parts of objects. The ensemble effects are obtained from a single network by selecting random hidden unit pairs, which means that a variety of localization maps are generated from a single image. Our approach does not require any additional training steps and only adds a simple layer to a standard convolutional neural network; nevertheless it outperforms recent comparable techniques on the Pascal VOC 2012 benchmark in both weakly and semi-supervised settings.

This review aims to define the 4 types of the metaverse and to explain the potential and limitations of its educational applications. The metaverse roadmap categorizes the metaverse into 4 types: augmented reality, lifelogging, mirror world, and virtual reality. An example of the application of augmented reality in medical education would be an augmented reality T-shirt that allows students to examine the inside of the human body as an anatomy lab. Furthermore, a research team in a hospital in Seoul developed a spinal surgery platform that applied augmented reality technology. The potential of the metaverse as a new educational environment is suggested to be as follows: a space for new social communication; a higher degree of freedom to create and share; and the provision of new experiences and high immersion through virtualization. Some of its limitations may be weaker social connections and the possibility of privacy impingement; the commission of various crimes due to the virtual space and anonymity of the metaverse; and maladaptation to the real world for students whose identity has not been established. The metaverse is predicted to change our daily life and economy beyond the realm of games and entertainment. The metaverse has infinite potential as a new social communication space. The following future tasks are suggested for the educational use of the metaverse: first, teachers should carefully analyze how students understand the metaverse; second, teachers should design classes for students to solve problems or perform projects cooperatively and creatively; third, educational metaverse platforms should be developed that prevent misuse of student data.

A variety of organisms, such as bacteria, fungi, and plants, produce secondary metabolites, also known as natural products. Natural products have been a prolific source and an inspiration for numerous medical agents with widely divergent chemical structures and biological activities, including antimicrobial, immunosuppressive, anticancer, and anti-inflammatory activities, many of which have been developed as treatments and have potential therapeutic applications for human diseases. Aside from natural products, the recent development of recombinant DNA technology has sparked the development of a wide array of biopharmaceutical products, such as recombinant proteins, offering significant advances in treating a broad spectrum of medical illnesses and conditions. Herein, we will introduce the structures and diverse biological activities of natural products and recombinant proteins that have been exploited as valuable molecules in medicine, agriculture and insect control. In addition, we will explore past and ongoing efforts along with achievements in the development of robust and promising microorganisms as cell factories to produce biologically active molecules. Furthermore, we will review multi-disciplinary and comprehensive engineering approaches directed at improving yields of microbial production of natural products and proteins and generating novel molecules. Throughout this article, we will suggest ways in which microbial-derived biologically active molecular entities and their analogs could continue to inspire the development of new therapeutic agents in academia and industry.