Search

Lianfeng Lu,1 Xiaodi Li,1 Xiaosheng Liu,2 Yang Han,1 Zhifeng Qiu,1 Xiaojing Song,1 Yanling Li,1 Xiaoxia Li,1 Wei Cao,1 Taisheng Li1,2 1Department of Infectious Diseases, Peking Union Medical College Hospital, Peking Union Medical College and Chinese Academy of Medical Sciences, Beijing, People’s Republic of China; 2Tsinghua-Peking Center for Life Sciences, Beijing, China Department of Basic Medical Sciences, School of Medicine, Tsinghua University, Beijing, People’s Republic of ChinaCorrespondence: Taisheng Li, Department of Infectious Diseases, Peking Union Medical College Hospital, Peking Union Medical College and Chinese Academy of Medical Sciences, 1# Shuai Fu Yuan, Beijing, 100730, People’s Republic of China, Email litsh@263.netObjective(s): HIV-1-infected Chinese patients who were treated naïve with combination dolutegravir (DTG) and tenofovir disoproxil fumarate (TDF) group, DTG without TDF group, TDF without DTG, as well as patients switched to DTG-containing therapy from other drugs were included.Design: The dynamics of serum creatinine, cystatin C (CysC) level, eGFRcr and eGFRCysC at the baseline, 4 w, 12w, 24w, 36w and 48w for different group of patients were collected and evaluated.Methods: Changes in serum creatinine, levels, eGFRcr and eGFRCysC were analyzed among groups and in different time-points. Intra-group correlation coefficient and Bland–Altman plot were used to compare the results of eGFRcr and eGFRCysC.Results: Thirty-seven treated-naïve HIV-patients in combined DTG and TDF group (group 1), 23 in DTG without TDF patients (group 2) and 47 patients on TDF without DTG group (control group, group 3) along with 31 patients whose ART switch to DTG-containing regimens (group 4) were collected. Serum creatinine was significantly elevated in the group 1 and group 2 instead of group 3 from baseline to 48w. Mean decreased change of eGFR calculated by serum creatinine proved the same conclusion. However, there were no differences in serum cysta

Latent Dirichlet allocation (LDA) is a widely-used probabilistic topic modeling tool for content analysis such as web mining. To handle web-scale content analysis on just a single PC, we propose multi-core parallel expectation-maximization (PEM) algorithms to infer and estimate LDA parameters in shared memory systems. By avoiding memory access conflicts reducing the locking time among multiple threads and residual-based dynamic scheduling, we show that PEM algorithms are more scalable and accurate than the current state-of-The-Art parallel LDA algorithms on a commodity PC. This parallel LDA toolbox is made publicly available as open source software at mloss.org.

Latent Dirichlet allocation (LDA) is a widely-used probabilistic topic modeling tool for content analysis such as web mining. To handle web-scale content analysis on just a single PC, we propose multi-core parallel expectation-maximization (PEM) algorithms to infer and estimate LDA parameters in shared memory systems. By avoiding memory access conflicts reducing the locking time among multiple threads and residual-based dynamic scheduling, we show that PEM algorithms are more scalable and accurate than the current state-of-The-Art parallel LDA algorithms on a commodity PC. This parallel LDA toolbox is made publicly available as open source software at mloss.org.

Latent Dirichlet allocation (LDA) is a widely-used probabilistic topic modeling tool for content analysis such as web mining. To handle web-scale content analysis on just a single PC, we propose multi-core parallel expectation-maximization (PEM) algorithms to infer and estimate LDA parameters in shared memory systems. By avoiding memory access conflicts reducing the locking time among multiple threads and residual-based dynamic scheduling, we show that PEM algorithms are more scalable and accurate than the current state-of-The-Art parallel LDA algorithms on a commodity PC. This parallel LDA toolbox is made publicly available as open source software at mloss.org.