The advent of advanced metagenome sequencing techniques has enabled the direct study of microbial communities in their natural habitats, thereby facilitating the acquisition of crucial insights into their composition and structure. As sequencing technology has advanced, the quality and length of sequences have continuously improved, making long-read segments and contigs more similar. Consequently, long-read segments can now be binned before assembly. However, a limited number of long-read binning methods exist, and some rely directly on composition and inaccurate coverage information. Additionally, while some existing methods attempt to improve the accuracy of their coverage estimation for individual long reads, their strategies remain imprecise, and fusion strategies of composition and coverage information also show unreliability. In this paper, we present improving metagenomic binning of long reads via feature fusion and Graph SAGE named as MBLR-GS, a scalable reference-free binning method. Specifically, it employs a new coverage estimation strategy and a feature fusion strategy that first enables fast and accurate estimation of coverage information using an overlap graph and utilizes a new feature fusion strategy that leverages composition and coverage information for binning during the binning process. Moreover, MBLR-GS incorporates sequence overlap information into the binning process using Graph SAGE. The proposed MBLR-GS outperforms existing methods in terms of binning accuracy on both simulated and real datasets, with an average improvement of about 6.32% in F1 score and 13.55% in ARI. In addition, MBLR-GS improves the accuracy without increasing the computation time. Furthermore, our method has been demonstrated to significantly improve binning accuracy without increasing computation time, thus facilitating subsequent downstream analyses.

We develop a high-performance tensor-based simulator for random quantum circuits(RQCs) on the new Sunway supercomputer. Our major innovations include: (1) a near-optimal slicing scheme, and a path-optimization strategy that considers both complexity and compute density; (2) a three-level parallelization scheme that scales to about 42 million cores; (3) a fused permutation and multiplication design that improves the compute efficiency for a wide range of tensor contraction scenarios; and (4) a mixed-precision scheme to further improve the performance. Our simulator effectively expands the scope of simulatable RQCs to include the 10X10(qubits)X(1+40+1)(depth) circuit, with a sustained performance of 1.2 Eflops (single-precision), or 4.4 Eflops (mixed-precision)as a new milestone for classical simulation of quantum circuits; and reduces the simulation sampling time of Google Sycamore to 304 seconds, from the previously claimed 10,000 years.

We present a deep learning-based thumbnail generation method called CropNet in this paper. Unlike previous deep learning-based methods, such as Fast-AT, which can utilize detectors introduced in object detection frameworks and generate thousands of proposals, our detector is straightforward and concise, thereby ensuring that the final cropping window is computed by its center and width, with the input aspect ratio. To achieve this goal, CropNet learns specific filters to estimate the center position and utilizes a cascade structure of filters and single neuron for width inference. In addition, CropNet optimizes the center and width jointly for optimal results. We collect a data set of more than 29,000 thumbnail annotations to train CropNet and perform cross-validation between existing data sets. Experiments show that CropNet outperforms existing techniques. Our result is achieved at a test-time speed of 17 ms per image, which is six times faster than the fastest method at present.

We present an automatic thumbnail generation technique based on two essential considerations: how well they visually represent the original photograph, and how well the foreground can be recognized after the cropping and downsizing steps of thumbnailing. These factors, while important for the image indexing purpose of thumbnails, have largely been ignored in previous methods, which instead are designed to highlight salient content while disregarding the effects of downsizing. We propose a set of image features for modeling these two considerations of thumbnails, and learn how to balance their relative effects on thumbnail generation through training on image pairs composed of photographs and their corresponding thumbnails created by an expert photographer. Experiments show the effectiveness of this approach on a variety of images, as well as its advantages over related techniques.