Circular RNAs are a special type of RNAs which recently attracted a lot of research interest in studying its formation and function. RNA binding proteins (RBPs) that bind circRNAs are important in these processes but are relatively less studied. CLIP-Seq technology has been invented and applied to profile RBP-RNA interactions on the genome-wide scale. While mRNAs are usually the focus of CLIP-Seq experiments, RBP-circRNA interactions could also be identified through specialized analysis of CLIP-Seq datasets. However, many technical difficulties are involved in this process, such as the usually short read length of CLIP-Seq reads. In this study, we created a pipeline called Clirc specialized for profiling circRNAs in CLIP-Seq data and analyzing the characteristics of RBP- circRNAs interactions. In conclusion, this is one of the first few studies to investigate circRNAs and their binding partners through repurposing CLIP-Seq datasets to our knowledge, and we hope our work will become a valuable resource for future studies into the biogenesis and function of circRNAs. Clirc software is available at https://github.com/Minzhe/Clirc

Domain generalization has led to remarkable achievements in Hyperspectral Image (HSI) classification. Inspired by contrastive language-image pre-training (CLIP), the language-aware domain generalization method has been explored to learn cross-domain-invariant representation. However, existing methods face some challenges: 1) The weak capacity to extract long-range contextual information and inter-class correlation. 2) Due to the inadequacies of the large-scale pre-training for HSI data, the spatial-spectral features of HSI and linguistic features can not be straightforwardly alignment. To address the above problems, a novel network has been proposed with a CLIP framework, which consists of an image encoder, based on an encoder-only transformer to obtain the global contextual information and inter-class correlation, a frozen text encoder, and a cross-attention mechanism, named Linguistic-Interact-with-Visual Engager (LIVE), enhances the interaction between two modalities. Extensive experiments demonstrating superior performance over state-of-the-art methods in HSI Domain Generalization with a CLIP framework.

Background and aims. Clinically significant delayed bleeding (CSDB) may complicate endoscopic colorectal submucosal dissection (ESD). Methods. We assessed the results of a prospective registry of colorectal ESD. We evaluated the effect of clip closure and PuraStat application for prevention of CSDB. Results. 40 patients with 41 colorectal ESD were included. ESD was successful in 38 lesions (92.7%), 35 with R0 resection (92.1%) and 33 with curative resection (86.8%). CSDB occurred in 3 out 38 lesions (7.9%, 95% CI [1.7 – 21.4%]), exclusively after rectal ESD (3 of 22 rectal lesions vs. 0 of 16 colonic lesions, p = 0.249). Clip closure was more frequently used after colonic ESD (12 of 16 colonic lesions vs. 2 of 22 rectal lesions, p < 0.001) and was not protective for CSDB in univariate analysis, even though no events occurred after clip closure (0 of 14 lesions with clip closure vs. 3 of 27 lesions without, p = 0.283). PuraStat was more frequently applied after ESD for rectal lesions (16 of 22 rectal lesions vs. 2 of 16 colonic le-sions, p < 0.001) and was not protective for CSDB, all three events occurring after PuraStat application (3 of 21 lesions with PuraStat application vs. 0 of 20 lesions without, p = 0.097). Conclusions. CSDB occurred exclusively after rectal ESD, and no predictive factors were identified in uni-variate analysis. Clip closure and PuraStat application were not protective for CSDB.

Visualising textual content could be helpful to professionals as well as amateurs across several fields. However, training a text-to-image generator in the mainstream domain requires large amounts of paired text-image and data, which is too expensive to collect since labeling millions of images and videos can be tiresome. GANs like StackGAN and StyleGAN can be considered as solutions to generate images from text. But the images generated may be of low accuracy and resolution, and the entire processing can be highly time-consuming. Moreover, image generation is a notion that is still being researched. Hence, the process of developing a Video Generation model necessitates substantial research. Despite the need for such a model, modern technology has lagged behind the solutions to this problem. This proposal suggests combining two methods, Text modification for Action Definition (TexAD) and SeQuential Image Generation for Video Synthesis (SQIGen). The proposed solution synthesises a sequence of images from textual information feeds and combines these images to create a video. TexAD uses Natural Language Processing and Deep Learning techniques to process, classify and modify text data. SQIGen is an extension of the VQGAN+CLIP neural network architecture that generates a sequence of images from the modified text data