Geometric graph learning with extended atom-types features for protein-ligand binding affinity prediction

Micronutrients have significant functional implications for the human immune response, and the quality of food is a major factor affecting the severity and mortality caused by HIV in individuals undergoing antiretroviral therapy. A decrease in CD4 lymphocyte count and an increase in CD8 lymphocyte count are the hallmarks of HIV infection, which causes the CD4/CD8 ratio to invert from a normal value of >1.6 to <1.0. In this study, we tried to analyze whether the nutritional status of HIV-positive patients has an impact on the CD4/CD8 ratio inversion by utilizing a machine learning (ML) algorithm. In this study, 55 confirmed HIV-positive patients who had not started their anti-retroviral therapy were included after obtaining their informed, written consent. Moreover, 55 age-and sex-matched relatives and caregivers of the patients who tested negative in the screening were enrolled as controls. All individual patient data points were analyzed for model development with an 80-20 train-test split. Four trace elements, zinc (Zn), phosphate (P), magnesium (Mg), and calcium (Ca), were utilized by implementing a random forest classifier. The target of the study was the inverted CD4/CD8 ratio. The data of 110 participants were included in the analysis. The algorithm thus generated had a sensitivity of 80% and a specificity of 83%, with a likelihood ratio (LR+) of 4.8 and LR-of 0.24. The utilization of the ML algorithm adds to the limited evidence that currently exists regarding the role of micronutrients, especially trace elements, in the causation of immune risk. Our inherent strength lies in the fact that this study is one of the first studies to utilize an ML-based decision tree algorithm to classify immune risk in HIV patients. Our study uniquely corroborated the nutritional data to the immune risk in treatment-naïve HIV patients through the utilization of a decision tree ML algorithm. This could subsequently be an important classification and prognostic tool in the hands of clinicians.

Sea ice seriously endangers the safe navigation of ships at sea. Therefore, it is the key to accurately judge the type of sea ice around the ship. However, remote sensing technology will be restricted by external weather changes. Therefore, this paper presents a method based on statistical feature extraction and machine learning of ship main engine vibration signal, so as to realize the recognition of sea ice types. Firstly, according to the sea ice video information captured by the camera, the acceleration and vibration data collected by the vibration sensor are divided, and the wavelet threshold denoising algorithm is adopted to enhance the features. Then, based on the statistical characteristics of vibration signals, time domain and frequency domain features are extracted. Finally, Random Forest(RF) and Gradient Boosting Decision Tree(GBDT) machine learning algorithms are used to classify the extracted eigenvalues. The experimental results show that the accuracy of sea ice identification using ship main engine vibration signals is 79%. This method provides a reference for identifying sea ice types by ship ice-induced vibration, and is helpful for further research on sea ice classification.

Right heart failure (RHF) remains a common and serious complication after durable left ventricular assist device (LVAD) implantation. We used explainable machine learning (ML) methods to derive novel insights into preimplant patient factors associated with RHF. Continuous-flow LVAD implantations from 2008 to 2017 in the Interagency Registry for Mechanically Assisted Circulatory Support (INTERMACS) were included. A total of 186 preimplant patient factors were analyzed and the primary outcome was 30 days of severe RHF. A boosted decision tree ML algorithm and an explainable ML method were applied to identify the most important factors associated with RHF, nonlinear relationships and interactions, and risk inflection points. Out of 19,595 patients, 19.1% developed severe RHF at 30 days. Thirty top predictors of RHF were identified with the top five being INTERMACS profile, Model for End-stage Liver Disease score, the number of inotropic infusions, hemoglobin, and race. Many top factors exhibited nonlinear relationships with key risk inflection points such as INTERMACS profile between 2 and 3, right atrial pressure of 15 mmHg, pulmonary artery pressure index of 3, and prealbumin of 23 mg/dl. Finally, the most important variable interactions involved INTERMACS profile and the number of inotropes. These insights could help formulate patient optimization strategies prior to LVAD implantation.

Text mining is a special case of data mining which explore unstructured or semi-structured text documents, to establish valuable patterns and rules that indicate trends and significant features about specific topics. Text mining has been in pattern recognition, predictive studies, sentiment analysis and statistical theories in many areas of research, medicine, financial analysis, social life analysis, and business intelligence. Text mining uses concept of natural language processing and machine learning. Machine learning algorithms have been used and reported to give great results, but their performance of machine learning algorithms is affected by factors such as dataset domain, number of classes, length of the corpus, and feature selection techniques used. Redundant attribute affects the performance of the classification algorithm, but this can be reduced by using different feature selection techniques and dimensionality reduction techniques. Feature selection is a data preprocessing step that chooses a subset of input variable while eliminating features with little or no predictive information. Feature selection techniques are Information gain, Term Frequency, Term Frequency-Inverse document frequency, Mutual Information, and Chi-Square, which can use a filters, wrappers, or embedded approaches. To get the most value from machine learning, pairing the best algorithms with the right tools and processes is necessary. Little research has been done on the effect of feature selection techniques on classification accuracy for pairing of these algorithms with the best feature selection techniques for optimal results. In this research, a text classification experiment was conducted using incident management dataset, where incidents were classified into their resolver groups. Support vector machine (SVM), K-Nearest Neighbors (KNN), Naïve Bayes (NB) and Decision tree (DT) machine learning algorithms were examined. Filtering approach was used on the feature selection techniques, with different ranking indices applied for optimal feature set and classification accuracy results analyzed. The classification accuracy results obtained using TF were, 88% for SVM, 70% for NB, 79% for Decision tree, and KNN had 55%, while Boolean registered 90%, 83%, 82% and 75%, for SVM, NB, DT, and KNN respectively. TF-IDF, had 91%, 83%, 76%, and 56% for SVM, NB, DT, and KNN respectively. The results showed that algorithm performance is affected by feature selection technique applied. SVM performed best, followed by DT, KNN and finally NB. In conclusion, presence of noisy data leads to poor learning performance and increases the computational time. The classifiers performed differently depending on the feature selection technique applied. For optimal results, the classifier that performed best together with the feature selection technique with the best feature subset should be applied for all types of data for accurate classification performance. Keywords: Text Classification, Supervised Machine Learning, Feature Selection DOI : 10.7176/JIEA/9-3-06 Publication date :May 31 st 2019

Rainfall prediction is the highest research priority in flood-prone areas across the world. This work assesses the abilities of the Decision Tree (DT), Distributed Decision Tree (DDT), Naïve Bayes (NB), Random Forest (RF), Support Vector Machine (SVM), K Nearest Neighbour (KNN), and Fuzzy Logic Decision Tree (FDTs) machine learning algorithms for the rainfall prediction across the Kashmir province of the Union Territory of Jammu & Kashmir. On application of Machine learning algorithms on geographical datasets gave performance accuracy varying from (78.61–81.53)%. Further again machine learning algorithms were reapplied on the dataset without season variable yet again performance ranged in between (77.5–81)%. Vigorous analysis has established that these machine learning models are robust and our study has established that the dataset reaches performance stagnation and thus resulting in performance capping. The stagnation is irrespective of the choice of algorithm and the performance shall not improvise beyond a specific value irrespective of the choice of the machine learning algorithm.KeywordsDecision treeGeographical dataRandom forestNaïve BayesDistribute decision tree

Code smells are generally not considered as bugs; instead, they point out certain shortcomings in the software design or code. Identification of code smell is a necessary step for improving the software quality and reducing the maintenance effort. In this study, we introduce a bad smell prediction technique based on object-oriented software metrics that use Decision Tree (DT) and Random Forest (RF) machine learning algorithm. An open-source project, namely JHOTDRAW, was used as our dataset, for which values of object-oriented software metrics were calculated. Two feature selection methods-Random Forest Importance (RFI) and Information Gain (IG) were applied to extract the most relevant attributes for the prediction of code smells, namely, Feature envy, Dispersed coupling, refused parent bequest, and God class. The random-search algorithm was used to tune the parameters of Random Forest and Decision Tree. Results show that the best classification accuracy for Decision Tree was obtained at 99.13% for refused parent bequest code smells. Results also show that after using the Random Forest classifier, refused parent bequest smell was predicted with the best accuracy of 99.14%. Finally, in this research study, a set of code smell prediction rules were extracted using Decision Tree.

Genetic testing can determine family screening strategies and has prognostic and diagnostic value in hypertrophic cardiomyopathy (HCM). However, it can also pose a significant psychosocial burden. Conventional scoring systems offer modest ability to predict genotype positivity. The aim of our study was to develop a novel prediction model for genotype positivity in patients with HCM by applying machine learning (ML) algorithms. We constructed 3 ML models using readily available clinical and cardiac imaging data of 102 patients from Columbia University with HCM who had undergone genetic testing (the training set). We validated model performance on 76 patients with HCM from Massachusetts General Hospital (the test set). Within the test set, we compared the area under the receiver operating characteristic curves (AUROCs) for the ML models against the AUROCs generated by the Toronto HCM Genotype Score (the Toronto score) and Mayo HCM Genotype Predictor (the Mayo score) using the Delong test and net reclassification improvement. Overall, 63 of the 178 patients (35%) were genotype positive. The random forest ML model developed in the training set demonstrated an AUROC of 0.92 (95% CI, 0.85-0.99) in predicting genotype positivity in the test set, significantly outperforming the Toronto score (AUROC, 0.77 [95% CI, 0.65-0.90], P=0.004, net reclassification improvement: P<0.001) and the Mayo score (AUROC, 0.79 [95% CI, 0.67-0.92], P=0.01, net reclassification improvement: P=0.001). The gradient boosted decision tree ML model also achieved significant net reclassification improvement over the Toronto score (P<0.001) and the Mayo score (P=0.03), with an AUROC of 0.87 (95% CI, 0.75-0.99). Compared with the Toronto and Mayo scores, all 3 ML models had higher sensitivity, positive predictive value, and negative predictive value. Our ML models demonstrated a superior ability to predict genotype positivity in patients with HCM compared with conventional scoring systems in an external validation test set.

Protein-ligand interactions are the molecular basis of many important cellular activities, such as gene regulation, cell metabolism, and signal transduction. Protein-ligand binding affinity is a crucial metric of the strength of the interaction between the two, and accurate prediction of its binding affinity is essential for discovering drugs' new uses. So far, although many predictive models based on machine learning and deep learning have been reported, most of the models mainly focus on one-dimensional sequence and two-dimensional structural characteristics of proteins and ligands, but fail to deeply explore the detailed interaction information between proteins and ligand atoms in the binding pocket region of three-dimensional space. In this study, we introduced a novel 4D tensor feature to capture key interactions within the binding pocket and developed a three-dimensional convolutional neural network (CNN) model based on this feature. Using ten-fold cross-validation, we identified the optimal parameter combination and pocket size. Additionally, we employed feature engineering to extract features across multiple dimensions, including one-dimensional sequences, two-dimensional structures of the ligand and protein, and three-dimensional interaction features between them. We proposed an efficient protein-ligand binding affinity prediction model MCDTA (multi-dimensional convolutional drug-target affinity), built on a multi-dimensional convolutional neural network framework. Feature ablation experiments revealed that the 4D tensor feature had the most significant impact on model performance. MCDTA performed exceptionally well on the PDBbind v.2020 dataset, achieving an RMSE of 1.231 and a PCC of 0.823. In comparative experiments, it outperformed five other mainstream binding affinity prediction models, with an RMSE of 1.349 and a PCC of 0.795. Moreover, MCDTA demonstrated strong generalization ability and practical screening performance across multiple benchmark datasets, highlighting its reliability and accuracy in predicting protein-ligand binding affinity. The code for MCDTA is available at https://github.com/dfhuang-AI/MCDTA .

Predicting protein-ligand binding affinity is a central issue in drug design. Various deep learning models have been published in recent years, where many of them rely on 3D protein-ligand complex structures as input and tend to focus on the single task of reproducing binding affinity. In this study, we have developed a graph neural network model called PLANET (Protein-Ligand Affinity prediction NETwork). This model takes the graph-represented 3D structure of the binding pocket on the target protein and the 2D chemical structure of the ligand molecule as input. It was trained through a multi-objective process with three related tasks, including deriving the protein-ligand binding affinity, protein-ligand contact map, and ligand distance matrix. Besides the protein-ligand complexes with known binding affinity data retrieved from the PDBbind database, a large number of non-binder decoys were also added to the training data for deriving the final model of PLANET. When tested on the CASF-2016 benchmark, PLANET exhibited a scoring power comparable to the best result yielded by other deep learning models as well as a reasonable ranking power and docking power. In virtual screening trials conducted on the DUD-E benchmark, PLANET's performance was notably better than several deep learning and machine learning models. As on the LIT-PCBA benchmark, PLANET achieved comparable accuracy as the conventional docking program Glide, but it only spent less than 1% of Glide's computation time to finish the same job because PLANET did not need exhaustive conformational sampling. Considering the decent accuracy and efficiency of PLANET in binding affinity prediction, it may become a useful tool for conducting large-scale virtual screening.

The design of minimum CRISPR RNA (crRNA) sets for detection of diverse RNA targets using sequence degeneracy has not been systematically addressed. We tested candidate degenerate Cas13a crRNA sets designed for detection of diverse RNA targets (Lassa virus). A decision tree machine learning (ML) algorithm (RuleFit) was applied to define the top attributes that determine the specificity of degenerate crRNAs to elicit collateral nuclease activity. Although the total number of mismatches (0–4) is important, the specificity depends as well on the spacing of mismatches, and their proximity to the 5’ end of the spacer. We developed a predictive algorithm for design of candidate degenerate crRNA sets, allowing improved discrimination between “included” and “excluded” groups of related target sequences. A single degenerate crRNA set adhering to these rules detected representatives of all Lassa lineages. Our general ML approach may be applied to the design of degenerate crRNA sets for any CRISPR/Cas system.

In the drug discovery process and repurposing of existing drugs, accurately identifying ligands with high binding affinity to proteins is a very critical step. However, it sinks a lot of time and resources to detect the protein-ligand binding affinity through biological experiments. Therefore, it is very necessary to develop an accurate and reliable computational method to predict the binding affinity between protein and ligand. At present, some computational methods have been proposed to predict the protein-ligand binding affinity, but the absence of protein-ligand complexes structures restricts some predictive methods that require input the complexes structures. In this paper, a novel deep-learning-based method is proposed, named DPLA, to predict binding affinity by integrating multilevel information of protein and ligand. More specifically, our model extracted some important information, such as sequence representation, structural property representation of amino acids in protein and protein binding pocket, MACCS key ligand molecular fingerprint and ligand molecular network features. This method was tested on the PDBbind core set, and we compared it with some recent state-of-art protein-ligand affinity prediction methods. The excellent performance shows that DPLA is an accurate and reliable method for affinity prediction.

Accurately predicting protein-ligand binding affinity is an important step in the drug discovery process. Deep learning (DL) methods have improved binding affinity prediction by using diverse categories of molecular data. However, many models rely heavily on interaction and sequence data, which impedes proper learning and limits performance in de novo applications. To address these limitations, we developed a novel graph neural network model, called StructureNet (structure-based graph neural network), to predict protein-ligand binding affinity. StructureNet improves existing DL methods by focusing entirely on structural descriptors to mitigate data memorization issues introduced by sequence and interaction data. StructureNet represents the protein and ligand structures as graphs, which are processed using a GNN-based ensemble deep learning model. StructureNet achieved a PCC of 0.68 and an AUC of 0.75 on the PDBBind v.2020 Refined Set, outperforming similar structure-based models. External validation on the DUDE-Z dataset showed that StructureNet can effectively distinguish between active and decoy ligands. Further testing on a small subset of well-known drugs indicates that StructureNet has high potential for rapid virtual screening applications. We also hybridized StructureNet with interaction- and sequence-based models to investigate their impact on testing accuracy and found minimal difference (0.01 PCC) between merged models and StructureNet as a standalone model. An ablation study found that geometric descriptors were the key drivers of model performance, with their removal leading to a PCC decrease of over 15.7%. Lastly, we tested StructureNet on ensembles of binding complex conformers generated using molecular dynamics (MD) simulations and found that incorporating multiple conformations of the same complex often improves model accuracy by capturing binding site flexibility. Overall, the results show that structural data alone are sufficient for binding affinity predictions and can address pattern recognition challenges introduced by sequence and interaction features. Additionally, structural representations of protein-ligand complexes can be considerably improved using geometric and topological descriptors. We made StructureNet GUI interface freely available online.

With the accumulation of smoking data and the development of the algorithm, precise analysis become possible and this can benefit smoking cessation a lot. However, as far as we know, little research has been done on the behavior of everyday smoking, such as the precise time when a smoker smokes. This paper proposes a model based on decision tree machine learning algorithm to predict daily smoking time. The simulation data set of smoking time data was established by using the population information of smokers collected by the Chinese center for disease control and prevention. In order to solve the problem of too little feature information, we propose a feature information extraction module. In this paper, we tested a variety of machine learning algorithms, and finally came to the conclusion that the prediction model based on XGBoost had the best performance, with an accuracy rate of 84.11%, and its training was much faster than that based on other machine learning algorithms

This paper presents a posture recognition system aimed at detecting sitting postures of a wheelchair user. The main goals of the proposed system are to identify and inform irregular and improper posture to prevent sitting-related health issues such as pressure ulcers, with the potential that it could also be used for individuals without mobility issues. In the proposed monitoring system, an array of 16 screen printed pressure sensor units was employed to obtain pressure data, which are sampled and processed in real-time using read-out electronics. The posture recognition was performed for four sitting positions: right-, left-, forward- and backward leaning based on k-nearest neighbors (k-NN), support vector machines (SVM), random forest (RF), decision tree (DT) and LightGBM machine learning algorithms. As a result, a posture classification accuracy of up to 99.03 percent can be achieved. Experimental studies illustrate that the system can provide real-time pressure distribution value in the form of a pressure map on a standard PC and also on a raspberry pi system equipped with a touchscreen monitor. The stored pressure distribution data can later be shared with healthcare professionals so that abnormalities in sitting patterns can be identified by employing a post-processing unit. The proposed system could be used for risk assessments related to pressure ulcers. It may be served as a benchmark by recording and identifying individuals’ sitting patterns and the possibility of being realized as a lightweight portable health monitoring device.

Predicting protein-ligand binding affinity is a crucial and challenging task in structure-based drug discovery. With the accumulation of complex structures and binding affinity data, various machine-learning scoring functions, particularly those based on deep learning, have been developed for this task, exhibiting superiority over their traditional counterparts. A fusion model sequentially connecting a graph neural network (GNN) and a convolutional neural network (CNN) to predict protein-ligand binding affinity is proposed in this work. In this model, the intermediate outputs of the GNN layers, as supplementary descriptors of atomic chemical environments at different levels, are concatenated with the input features of CNN. The model demonstrates a noticeable improvement in performance on CASF-2016 benchmark compared to its constituent CNN models. The generalization ability of the model is evaluated by setting a series of thresholds for ligand extended-connectivity fingerprint similarity or protein sequence similarity between the training and test sets. Masking experiment reveals that model can capture key interaction regions. Furthermore, the fusion model is applied to a virtual screening task for a novel target, PI5P4Kα. The fusion strategy significantly improves the ability of the constituent CNN model to identify active compounds. This work offers a novel approach to enhancing the accuracy of deep learning models in predicting binding affinity through fusion strategies.