More Like this

1. Inter- and Intra- Patient ECG Heartbeat Classification for Arrhythmia Detection: A Sequence to Sequence Deep Learning Approach

   https://doi.org/10.1109/ICASSP.2019.8683140  

   Mousavi, Sajad ; Afghah, Fatemeh ( May 2019 , ICASSP 2019 - 2019 IEEE International Conference on Acoustics, Speech and Signal Processing (ICASSP))
2. Sequence and Nuclease Requirements for Breakage and Healing of a Structure-Forming (AT)n Sequence within Fragile Site FRA16D

   https://doi.org/10.1016/j.celrep.2019.03.103  

   Kaushal, Simran ; Wollmuth, Charles E. ; Das, Kohal ; Hile, Suzanne E. ; Regan, Samantha B. ; Barnes, Ryan P. ; Haouzi, Alice ; Lee, Soo Mi ; House, Nealia C.M. ; Guyumdzhyan, Michael ; et al ( April 2019 , Cell Reports)
3. A Reinforced Topic-Aware Convolutional Sequence-to-Sequence Model for Abstractive Text Summarization

   https://doi.org/10.24963/ijcai.2018/619  

   Wang, Li ; Yao, Junlin ; Tao, Yunzhe ; Zhong, Li ; Liu, Wei ; Du, Qiang ( July 2018 , International Joint Conference on Artificial Intelligence)

   In this paper, we propose a deep learning approach to tackle the automatic summarization tasks by incorporating topic information into the convolutional sequence-to-sequence (ConvS2S) model and using self-critical sequence training (SCST) for optimization. Through jointly attending to topics and word-level alignment, our approach can improve coherence, diversity, and informativeness of generated summaries via a biased probability generation mechanism. On the other hand, reinforcement training, like SCST, directly optimizes the proposed model with respect to the non-differentiable metric ROUGE, which also avoids the exposure bias during inference. We carry out the experimental evaluation with state-of-the-art methods over the Gigaword, DUC-2004, and LCSTS datasets. The empirical results demonstrate the superiority of our proposed method in the abstractive summarization.

    

   more » « less
4. 9.8 A 25mm 2 SoC for IoT Devices with 18ms Noise-Robust Speech-to-Text Latency via Bayesian Speech Denoising and Attention-Based Sequence-to-Sequence DNN Speech Recognition in 16nm FinFET

   https://doi.org/10.1109/ISSCC42613.2021.9366062  

   Tambe, Thierry ; Yang, En-Yu ; Ko, Glenn G. ; Chai, Yuji ; Hooper, Coleman ; Donato, Marco ; Whatmough, Paul N. ; Rush, Alexander M. ; Brooks, David ; Wei, Gu-Yeon ( February 2021 , International Solid-State Circuits Conference)
5. Comparative Analysis and Ancestral Sequence Reconstruction of Bacterial Sortase Family Proteins Generates Functional Ancestral Mutants with Different Sequence Specificities

   https://doi.org/10.3390/bacteria1020011  

   Valgardson, Jordan D. ; Struyvenberg, Sarah A. ; Sailer, Zachary R. ; Piper, Isabel M. ; Svendsen, Justin E. ; Johnson, D. Alex ; Vogel, Brandon A. ; Antos, John M. ; Harms, Michael J. ; Amacher, Jeanine F. ( June 2022 , Bacteria)

   Gram-positive bacteria are some of the earliest known life forms, diverging from gram-negative bacteria 2 billion years ago. These organisms utilize sortase enzymes to attach proteins to their peptidoglycan cell wall, a structural feature that distinguishes the two types of bacteria. The transpeptidase activity of sortases make them an important tool in protein engineering applications, e.g., in sortase-mediated ligations or sortagging. However, due to relatively low catalytic efficiency, there are ongoing efforts to create better sortase variants for these uses. Here, we use bioinformatics tools, principal component analysis and ancestral sequence reconstruction, in combination with protein biochemistry, to analyze natural sequence variation in these enzymes. Principal component analysis on the sortase superfamily distinguishes previously described classes and identifies regions of relatively high sequence variation in structurally-conserved loops within each sortase family, including those near the active site. Using ancestral sequence reconstruction, we determined sequences of ancestral Staphylococcus and Streptococcus Class A sortase proteins. Enzyme assays revealed that the ancestral Streptococcus enzyme is relatively active and shares similar sequence variation with other Class A Streptococcus sortases. Taken together, we highlight how natural sequence variation can be utilized to investigate this important protein family, arguing that these and similar techniques may be used to discover or design sortases with increased catalytic efficiency and/or selectivity for sortase-mediated ligation experiments. 

   more » « less