A5014081501- Cellular transport and secretion
- Nuclear Structure and Function
- RNA and protein synthesis mechanisms
- Reinforcement Learning in Robotics
- Robotic Path Planning Algorithms
- Genetic Neurodegenerative Diseases
- Erythrocyte Function and Pathophysiology
- Modular Robots and Swarm Intelligence
Rutgers, The State University of New Jersey
2024
Singapore Bioimaging Consortium
2012-2013
Agency for Science, Technology and Research
2012-2013
Heterozygosity for missense mutations (N88S/S90L) in BSCL2 (Berardinelli–Seip congenital lipodystrophy type 2)/Seipin is associated with a broad spectrum of motoneuron diseases. To understand the underlying mechanisms how lead to motor neuropathy, we generated transgenic mice neuron-specific expression wild-type (tgWT) or N88S/S90L mutant (tgMT) human Seipin. Transgenes led WT Seipin brain and spinal cord. TgMT, but not tgWT, exhibited late-onset altered locomotor activities gait...
Heterozygosity for missense mutations in Seipin, namely N88S and S90L, leads to a broad spectrum of motor neuropathy, while number loss-of-function Seipin are associated with the Berardinelli-Seip congenital generalized lipodystrophy type 2 (CGL2, BSCL2), condition that is characterized by severe lipoatrophy, insulin resistance, intellectual impairment. The mechanisms which lead lipodystrophy, role plays central nervous system (CNS) remain unknown. goal this study understand functions CNS...
Background While pathogenic mutations in BSCL2/Seipin cause congenital generalized lipodystrophy, the underlying mechanism is largely unknown. In this study, we investigated whether and how missense A212P mutation of Seipin (Seipin-A212P) inhibits adipogenesis. Methodology/Results We analyzed gene expression lipid accumulation stable 3T3-L1 cell lines expressing wild type (3T3-WT), non-lipodystrophic mutants N88S (3T3-N88S) S90L (3T3-S90L), or lipodystrophic mutant (3T3-A212P). When treated...
Deep reinforcement learning (RL) has led to encouraging successes in numerous challenging robotics applications. However, the lack of inductive biases support logic deduction and generalization representation a deep RL model causes it less effective exploring complex long-horizon robot-control tasks with sparse reward signals. Existing program synthesis algorithms for problems inherit same limitation, as they either adapt conventional guide search or synthesize programs imitate an model. We...