Novel QSPR modeling of stability constants of metal-thiosemicarbazone complexes by hybrid multivariate technique: GA-MLR, GA-SVR and GA-ANN

Novel Thiosemicarbazone Quantum Dots in the Treatment of Alzheimer’s Disease Combining In Silico Models Using Fingerprints and Physicochemical Descriptors

“…The functional dependence of the variables pIC 50 and BE on a set of molecular descriptors in the support vector regression model (QSAR SVR ) needs to be estimated. , Similar to other regression methods, the relationship between the molecular descriptors and the pIC 50 activities and BE values are given by a deterministic function f ( x ) plus the addition of some additive noise y = f ( x ) + normaln normalo normali normals normale …”

“…A function f ( x ) needs to be defined to allow accurate prediction of new cases. This can be achieved by training the QSAR SVR model on a training set. , Depending on the definition of the error function f ( x ), the function type can be identified. The first error function used is known as the epsilon-SVM regression 0.5 w normalT w + C ∑ i = 1 N ξ i + C ∑ i = 1 N ξ i * …”

“…Kernel radial basis function (RBF) can be chosen for constructing the Support Vector Machines Regression model. , Here, C is the capacity constant and should be chosen carefully to avoid overfitting, w is the vector of coefficients, ξ i represents the parameters for handling nonseparable data (inputs), and the i index labels the N training cases. It should be noted that the larger the C , the more the error is penalized. , K ( X i , X j ) = { exp false( prefix− γ false| X i − X j false| 2 false) } where K ( X i · X j ) = ϕ X i ·ϕ X j ; i.e., the kernel function represents a dot product of input data points mapped into the higher dimensional feature space by transformation ϕ. The kernel ϕ is used to transform data from the independent input to the feature space.…”

Multitask Neural Network for Mapping the Glass Transition and Melting Temperature Space of Homo- and Co-Polyhydroxyalkanoates Using σProfiles Molecular Inputs

“…ANN, a common algorithm for nonlinear ML, includes a considerable number of processing units known as neurons interlinked to produce a complex network architecture, with its inputs, outputs, and hidden layers being the most typical classifications for this structure. Compared to the conventional multilinear model, ANN attracts attention due to its adaptability to model highly nonlinear problems, which are commonly used to understand and predict the correlation between input parameters and outputs. − The first layer represents the input, with each neuron signifying a descriptive variable, whereas the second layer represents the hidden layer. This layer analyzes the data from the input layer using a transfer activation function.…”

Search for conformation of thiosemicarbazone reagents and their complexes with metals by using Monte Carlo and docking simulation

“…Another regimen is using chloroquine, and the world is now following this direction.These drugs are all in the process of evaluating and researching treatment for SARS-CoV-2 patients in Vietnam. There are currently also drugs with antimicrobial, antifungal, and antiviral 2+ , Cu 2+ , Hg 2+ , Mn 2+ , Ni 2+ , Pd 2+ , and Zn 2+ , designed and synthesized by Quang et al[17,18] are used to perform docking on the active site of the SARS-CoV-2 protein. We predicted the protein spatial structure of SARS-CoV-2 from a database.…”