More Like this

1. The Frequency and Topology of Pseudoorthologs

   https://doi.org/10.1093/sysbio/syab097  

   Smith, Megan L.; Hahn, Matthew W.; Smith, ed., Stephen (December 2021, Systematic Biology)

   Abstract Phylogenetics has long relied on the use of orthologs, or genes related through speciation events, to infer species relationships. However, identifying orthologs is difficult because gene duplication can obscure relationships among genes. Researchers have been particularly concerned with the insidious effects of pseudoorthologs—duplicated genes that are mistaken for orthologs because they are present in a single copy in each sampled species. Because gene tree topologies of pseudoorthologs may differ from the species tree topology, they have often been invoked as the cause of counterintuitive results in phylogenetics. Despite these perceived problems, no previous work has calculated the probabilities of pseudoortholog topologies or has been able to circumscribe the regions of parameter space in which pseudoorthologs are most likely to occur. Here, we introduce a model for calculating the probabilities and branch lengths of orthologs and pseudoorthologs, including concordant and discordant pseudoortholog topologies, on a rooted three-taxon species tree. We show that the probability of orthologs is high relative to the probability of pseudoorthologs across reasonable regions of parameter space. Furthermore, the probabilities of the two discordant topologies are equal and never exceed that of the concordant topology, generally being much lower. We describe the species tree topologies most prone to generating pseudoorthologs, finding that they are likely to present problems to phylogenetic inference irrespective of the presence of pseudoorthologs. Overall, our results suggest that pseudoorthologs are unlikely to mislead inferences of species relationships under the biological scenarios considered here.[Birth–death model; orthologs; paralogs; phylogenetics.] 

   more » « less
2. Measuring the frequency response of the honeybee thorax

   https://doi.org/10.1088/1748-3190/ab835b  

   Jankauski, Mark A (July 2020, Bioinspiration & Biomimetics)
3. Estimating the frequency of a clustered signal

   Chen, Xue; Price, Eric (July 2019, 46th International Colloquium on Automata, Languages and Programming (ICALP 2019))
4. Passive Stability of Stance is Determined by the Relationship Between Natural Frequency and Walking Frequency

   https://doi.org/10.1007/978-3-031-72597-5_30  

   Riddle, Shane; Sutton, Gregory; Webster-Wood, Victoria A; Chiel, Hillel J; Quinn, Roger D (December 2024, Springer Nature Switzerland)
5. Learning in the Frequency Domain

   https://doi.org/10.1109/CVPR42600.2020.00181  

   Xu, Kai; Qin, Minghai; Sun, Fei; Wang, Yuhao; Chen, Yen-Kuang; Ren, Fengbo (January 2020, IEEE Conference on Computer Vision and Pattern Recognition)

   null (Ed.)

   Deep neural networks have achieved remarkable success in computer vision tasks. Existing neural networks mainly operate in the spatial domain with fixed input sizes. For practical applications, images are usually large and have to be downsampled to the predetermined input size of neural networks. Even though the downsampling operations reduce computation and the required communication bandwidth, it removes both redundant and salient information obliviously, which results in accuracy degradation. Inspired by digital signal processing theories, we analyze the spectral bias from the frequency perspective and propose a learning-based frequency selection method to identify the trivial frequency components which can be removed without accuracy loss. The proposed method of learning in the frequency domain leverages identical structures of the well-known neural networks, such as ResNet-50, MobileNetV2, and Mask R-CNN, while accepting the frequency-domain information as the input. Experiment results show that learning in the frequency domain with static channel selection can achieve higher accuracy than the conventional spatial downsampling approach and meanwhile further reduce the input data size. Specifically for ImageNet classification with the same input size, the proposed method achieves 1.60% and 0.63% top-1 accuracy improvements on ResNet-50 and MobileNetV2, respectively. Even with half input size, the proposed method still improves the top-1 accuracy on ResNet-50 by 1.42%. In addition, we observe a 0.8% average precision improvement on Mask R-CNN for instance segmentation on the COCO dataset. 

   more » « less