Search for: All records

Ever-growing edge applications often require short processing latency and high energy efficiency to meet strict timing and power budget. In this work, we propose that the compact long short-term memory (LSTM) model can approximate conventional acausal algorithms with reduced latency and improved efficiency for real-time causal prediction, especially for the neural signal processing in closed-loop feedback applications. We design an LSTM inference accelerator by taking advantage of the fine-grained parallelism and pipelined feedforward and recurrent updates. We also propose a bit-sparse quantization method that can reduce the circuit area and power consumption by replacing the multipliers with the bit-shift operators. We explore different combinations of pruning and quantization methods for energy-efficient LSTM inference on datasets collected from the electroencephalogram (EEG) and calcium image processing applications. Evaluation results show that our proposed LSTM inference accelerator can achieve 1.19 GOPS/mW energy efficiency. The LSTM accelerator with 2-sbit/16-bit sparse quantization and 60% sparsity can reduce the circuit area and power consumption by 54.1% and 56.3%, respectively, compared with a 16-bit baseline implementation.

In recent years, the field of neuroscience has gone through rapid experimental advances and a significant increase in the use of quantitative and computational methods. This growth has created a need for clearer analyses of the theory and modeling approaches used in the field. This issue is particularly complex in neuroscience because the field studies phenomena that cross a wide range of scales and often require consideration at varying degrees of abstraction, from precise biophysical interactions to the computations they implement. We argue that a pragmatic perspective of science, in which descriptive, mechanistic, and normative models and theories each play a distinct role in defining and bridging levels of abstraction, will facilitate neuroscientific practice. This analysis leads to methodological suggestions, including selecting a level of abstraction that is appropriate for a given problem, identifying transfer functions to connect models and data, and the use of models themselves as a form of experiment.

Apicomplexan parasites like Toxoplasma gondii grow and replicate within a specialized organelle called the parasitophorous vacuole. The vacuole is decorated with parasite proteins that integrate into the membrane after trafficking through the parasite secretory system as soluble, chaperoned complexes. A regulator of this process is an atypical protein kinase called WNG1. Phosphorylation by WNG1 appears to serve as a switch for membrane integration. However, like its substrates, WNG1 is secreted from the parasite dense granules, and its activity must, therefore, be tightly regulated until the correct membrane is encountered. Here, we demonstrate that, while another member of the WNG family can adopt multiple multimeric states, WNG1 is monomeric and therefore not regulated by multimerization. Instead, we identify two phosphosites on WNG1 that are required for its kinase activity. Using a combination of in vitro biochemistry and structural modeling, we identify basic residues that are also required for WNG1 activity and appear to recognize the activating phosphosites. Among these coordinating residues are the ‘HRD’ Arg, which recognizes activation loop phosphorylation in canonical kinases. WNG1, however, is not phosphorylated on its activation loop, but rather on atypical phosphosites on its C-lobe. We propose a simple model in which WNG1 is activated bymore »increasing ATP concentration above a critical threshold once the kinase traffics to the parasitophorous vacuole.« less

Abstract Trace fossils record foraging behaviors, the search for resources in patchy environments, of animals in the rock record. Quantification of the strength, density, and nature of foraging behaviors enables the investigation of how these may have changed through time. Here, we present a novel approach to explore such patterns using spatial point process analyses to quantify the scale and strength of ichnofossil spatial distributions on horizontal bedding planes. To demonstrate the utility of this approach, we use two samples from the terminal Ediacaran Shibantan Member in South China (between 551 and 543 Ma) and the early Cambrian Nagaur Sandstone in northwestern India (between 539 and 509 Ma). We find that ichnotaxa on both surfaces exhibited significant nonhomogeneous lateral patterns, with distinct levels of heterogeneity exhibited by different types of trace fossils. In the Shibantan, two ichnotaxa show evidence for mutual positive aggregation over a shared resource, suggesting the ability to focus on optimal resource areas. Trace fossils from the Nagaur Sandstone exhibit more sophisticated foraging behavior, with greater niche differentiation. Critically, mark correlation functions highlight significant spatial autocorrelation of trace fossil orientations, demonstrating the greater ability of these Cambrian tracemakers to focus on optimal patches. Despite potential limitations, thesemore »analyses hint at changes in the development and optimization of foraging at the Ediacaran/Cambrian transition and highlight the potential of spatial point process analysis to tease apart subtle differences in behavior in the trace fossil record.« less

Pain relief on-demand Chronic pain is a debilitating condition for which there are no effective treatments. The primary somatosensory cortex (S1) and the anterior cingulate cortex (ACC) are involved in decoding pain components, and electrical stimulation of the prefrontal cortex (PFC) has been shown to exert analgesic effects. Here, Sun et al. developed a multiregion brain-machine interface (BMI) able to detect pain from electrical signals in S1 and ACC and provide on-demand PFC stimulation. The BMI was able to accurately detect and treat acute and chronic pain in rats; the analgesic effects were stable over time. The results suggest that BMI approaches might be effective for treating chronic pain of different etiologies.