Machine Learning for Sequence and Structure-Based Protein–Ligand Interaction Prediction
Y Zhang, S Li, K Meng, S Sun - Journal of Chemical Information …, 2024 - ACS Publications
… the prediction of the essential parameters, such as the binding … as well as the scoring functions
evaluated on scoring tasks. … nature of training samples, (2) lack of negative samples, and (…
evaluated on scoring tasks. … nature of training samples, (2) lack of negative samples, and (…
Assessing the information content of structural and protein–ligand interaction representations for the classification of kinase inhibitor binding modes via machine …
… scoring functions [13, 16]. However, IFPs might fail to detect key … to predict the complete
training set (ie 90% of the total data set). … for other hyper-parameters. Feature importance was …
training set (ie 90% of the total data set). … for other hyper-parameters. Feature importance was …
TB-IECS: an accurate machine learning-based scoring function for virtual screening
… in MLSFs and the characterization of protein–ligand interactions are always limited, which …
training process, SVM was used to find the best hyperplane to divide the positive and negative …
training process, SVM was used to find the best hyperplane to divide the positive and negative …
PIGNet: a physics-informed deep learning model toward generalized drug–target interaction predictions
… , in the comparative assessment of scoring functions (CASF) 2016, … physics of the
protein–ligand interaction as desired. For … can set a threshold for ΔG of a protein–ligand pair as …
protein–ligand interaction as desired. For … can set a threshold for ΔG of a protein–ligand pair as …
A High-Quality Data Set of Protein–Ligand Binding Interactions Via Comparative Complex Structure Modeling
X Li, C Shen, H Zhu, Y Yang, Q Wang… - Journal of Chemical …, 2024 - ACS Publications
… chemists to investigate protein–ligand interactions at the atomic level … We manually checked
any data points with a negative … the models trained on BindingNet could improve the scoring …
any data points with a negative … the models trained on BindingNet could improve the scoring …
Tapping on the black box: how is the scoring power of a machine-learning scoring function dependent on the training set?
… of scoring functions is crucial for distinguishing good and bad … the pairwise protein–ligand
interactions and protein–ligand … loss function was set to L2 loss, the penalty parameter C of …
interactions and protein–ligand … loss function was set to L2 loss, the penalty parameter C of …
Extended connectivity interaction features: improving binding affinity prediction through chemical description
… is its scoring function (SF), which is required to estimate the … of 9299 protein–ligand complexes
while the validation set … of the protein–ligand complexes denoted as pK, the negative base…
while the validation set … of the protein–ligand complexes denoted as pK, the negative base…
PSICHIC: physicochemical graph neural network for learning protein-ligand interaction fingerprints from sequence data
… Color gradient represents experimental potency (negative log-… score of 0.92 on the
protein-ligand functional effect test set. … -training stage, we categorized protein-ligand interactions …
protein-ligand functional effect test set. … -training stage, we categorized protein-ligand interactions …
New machine learning and physics-based scoring functions for drug discovery
… , lipophilic protein–ligand interactions and an improved estimation … Thus, we also trained
general scoring functions using all the … shown to be negative on MDM2 and Bcl2 interactions via …
general scoring functions using all the … shown to be negative on MDM2 and Bcl2 interactions via …
CarsiDock: a deep learning paradigm for accurate protein–ligand docking and screening based on large-scale pre-training
… structure information of protein–ligand interactions, but they … docking score was retained for
each protein–ligand pair, thus … the negative log-likelihood values of all protein–ligand atom …
each protein–ligand pair, thus … the negative log-likelihood values of all protein–ligand atom …