A new smoothing BFGS algorithm for constrained optimization problem

Yanxue Yang

Abstract: In this paper, the general constrained optimization problem is considered, the process of transforming it into unconstrained optimization problem by penalty function is given. And the smoothing function is used to smooth it, a smoothing BFGS algorithm is given to solve it. Under general conditions, the global convergence analysis of the algorithm is given, and the related numerical examples also show the effectiveness of the algorithm.

: Download (free)  CSE-PDF

QSAR Studies on Kinase Inhibitors of Fibroblast Growth Factor Receptor 4 Basis on Particle Swarm Optimization

Jize Wu, Junkang Liu, Xinyi Zhang, Honglin Zhai

Abstract: Hepatocellular carcinoma (HCC) is the sixth most prevalent cancer worldwide, typically developing in patients with underlying viral infections, metabolic disorders, or a history of alcohol abuse. Due to the variety adverse reactions to currently available treatments, there remains a clear requirement for agents that can provide improved outcomes in HCC. It is reported that fibroblast growth factor 19 (FGF19), signaling through fibroblast growth factor receptor 4 (FGFR4) and the coreceptor β-klotho (KLB), is implicated as the oncogenic driver in a subset of HCC, making selective FGFR4 inhibition an attractive treatment opportunity. In this work, quantitative structure-activity relationship (QSAR) studies were conducted on a series of FGFR4 inhibitors to shed light on the molecular requirements for inhibitor affinity and selectivity. Heuristic method (HM) is used to screen out five descriptor parameters, on this basis, multiple linear regression model was built. For the best linear model, the square of correlation coefficient (R2) was 0.67 and the square of cross validation correlation coefficient (R2CV) was 0.58. For nonlinear model, support vector machine (SVM) was used, with R2 and mean square error (S2) of 0.95 and 0.01 respectively for the training set, 0.84 and 0.04 respectively for the test set. It is apparent that the QSAR model based on SVM has better forecasting stability of inhibitor efficacy, and these findings should be useful for the design of FGFR4 inhibitors as highly selective first-in-class clinical candidate.

: Download (free)  CSE-PDF

Analysis of Enterprise Annuity in the Transportation System of a City in China

Fanrun Meng

Abstract: In 2004, China issued the Trial Measures for Enterprise Annuity, which initially established the enterprise annuity system and gradually became an important part of China's pension insurance system. Based on the investigation and analysis of the current situation of the participation of enterprise annuity in someone city's transportation system, this paper reveals the problems of the system in the aspects of personnel composition, operation organization, information disclosure, and supervision and management of the enterprise annuity. And put forward some corresponding suggestions to increase the insurance participation rate of enterprise annuity, and promote the sound and rapid development of enterprise annuity in transportation system.

: Download (free)  CSE-PDF