The emerging CRISPR/Cas9 system represents a promising platform for genome editing. However, its low transfection efficiency is a major problem hampering the application of the gene-editing potential of CRISPR/Cas9. Herein, by screening a pool of more than 56 kinds of agents, we constructed a novel polyethylene glycol phospholipid-modified cationic lipid nanoparticle (PLNP)-based delivery system that can condense and encapsulate a Cas9/single-guide RNA (sgRNA) plasmid (DNA) to form a core-shell structure (PLNP/DNA) that mediated up to 47.4% successful transfection of Cas9/sgPLK-1 plasmids in A375 cells in vitro. An intratumor injection of Cas9/sgPLK-1 plasmids into melanoma tumor-bearing mice resulted in significant downregulation of Pololike kinase 1 (PLK-1) protein and suppression of the tumor growth (467%) in vivo. This approach provides a versatile method that could be used for delivering the CRISPR/Cas9 system with high efficiency and safety both in vitro and in vivo.

The ability to quickly recognize and learn new visual concepts from limited samples enables humans to quickly adapt to new tasks and environments. This ability is enabled by semantic association of novel concepts with those that have already been learned and stored in memory. Computers can start to ascertain similar abilities by utilizing a semantic concept space. A concept space is a high-dimensional semantic space in which similar abstract concepts appear close and dissimilar ones far apart. In this paper, we propose a novel approach to one-shot learning that builds on this core idea. Our approach learns to map a novel sample instance to a concept, relates that concept to the existing ones in the concept space and, using these relationships, generates new instances, by interpolating among the concepts, to help learning. Instead of synthesizing new image instance, we propose to directly synthesize instance features by leveraging semantics using a novel auto-encoder network we call dual TriNet. The encoder part of the TriNet learns to map multi-layer visual features from CNN to a semantic vector. In semantic space, we search for related concepts, which are then projected back into the image feature spaces by the decoder portion of the TriNet. Two strategies in the semantic space are explored. Notably, this seemingly simple strategy results in complex augmented feature distributions in the image feature space, leading to substantially better performance.

Stable water-in-oil emulsion is essential to digital PCR and many other bioanalytical reactions that employ droplets as microreactors. We developed a novel technology to produce monodisperse emulsion droplets with high efficiency and high throughput using a bench-top centrifuge. Upon centrifugal spinning, the continuous aqueous phase is dispersed into monodisperse droplet jets in air through a micro-channel array (MiCA) and then submerged into oil as a stable emulsion. We performed dPCR reactions with a high dynamic range through the MiCA approach, and demonstrated that this cost-effective method not only eliminates the usage of complex microfluidic devices and control systems, but also greatly suppresses the loss of materials and cross-contamination. MiCA-enabled highly parallel emulsion generation combines both easiness and robustness of picoliter droplet production, and breaks the technical challenges by using conventional lab equipment and supplies.

Figure 1. Illustration of a variety of image deformations: ghosted (a, b), stitched (c), montaged (d), and partially occluded (e) images. AbstractHumans can robustly learn novel visual concepts even when images undergo various deformations and lose certain information. Mimicking the same behavior and synthesizing deformed instances of new concepts may help visual recognition systems perform better one-shot learning, i.e., learning concepts from one or few examples. Our key insight is that, while the deformed images may not be visually realistic, they still maintain critical semantic information and contribute significantly to formulating classifier decision boundaries. Inspired by the recent progress of meta-learning, we combine a meta-learner with an image deformation sub-network that produces additional training examples, and optimize both models in an end-to-end manner. The deformation sub-network learns to deform images by fusing a pair of images -a probe image that keeps the visual content and a gallery image that diversifies the deformations. We demonstrate results on the widely used one-shot learning benchmarks (miniImageNet and Im-ageNet 1K Challenge datasets), which significantly outperform state-of-the-art approaches. Code is available at https://github.com/tankche1/IDeMe-Net.