Average Matthews Correlation Coefficient Research Articles

Improving Protein-Protein Interaction Prediction Using Protein Language Model and Protein Network Features

Interactions between proteins are ubiquitous in a wide variety of biological processes. Accurately identifying the protein-protein interactions (PPI) is of significant importance for understanding the mechanisms of protein functions and facilitating drug discovery. Although the wet-lab technological methods are the best way to identify PPI, their major constraints are their time-consuming nature, high cost, and labor-intensiveness. Hence, lots of efforts have been made towards developing computational methods to improve the performance of PPI prediction. In this study, we propose a novel hybrid computational method (called KSGPPI) that aims at improving the prediction performance of PPI via extracting the discriminative information from protein sequences and interaction networks. The KSGPPI model comprises two feature extraction modules. In the first feature extraction module, a large protein language model, ESM-2, is employed to exploit the global complex patterns concealed within protein sequences. Subsequently, feature representations are further extracted through CKSAAP, and a two-dimensional convolutional neural network (CNN) is utilized to capture local information. In the second feature extraction module, the query protein acquires its similar protein from the STRING database via the sequence alignment tool NW-align and then captures the graph embedding feature for the query protein in the protein interaction network of the similar protein using the algorithm of Node2vec. Finally, the features of these two feature extraction modules are efficiently fused; the fused features are then fed into the fully connected neural networks to predict PPI. The results of five-fold cross-validation on the used benchmarked datasets demonstrate that KSGPPI achieves an average prediction accuracy of 88.96%. Additionally, the average Matthews correlation coefficient value (0.781) of KSGPPI is significantly higher than that of those state-of-the-art PPI prediction methods. The standalone package of KSGPPI is freely downloaded at https://github.com/rickleezhe/KSGPPI.

Sugar detection in adulterated honey using hyper-spectral imaging with stacking generalization method

This paper develops a new hybrid, automated, and non-invasive approach by combining hyper-spectral imaging, Savitzky–Golay (SG) Filter, Principal Components Analysis (PCA), Machine Learning (ML) classifiers/regressors, and stacking generalization methods to detect sugar in honey. First, the 32 different sugar concentration levels in honey were predicted using various ML regressors. Second, the six ranges of sugar were classified using various classifiers. Third, the 11 types of honey and 100% sugar were classified using classifiers. The stacking model (STM) obtained R2: 0.999, RMSE: 0.493 ml (v/v), RPD: 40.2, a 10-fold average R2: 0.996 and RMSE: 1.27 ml (v/v) for predicting 32 sugar concentrations. The STM achieved a Matthews Correlation Coefficient (MCC) of 99.7% and a Kappa score of 99.7%, a 10-fold average MCC of 98.9% and a Kappa score of 98.9% for classifying the six sugar ranges and 12 categories of honey types and a sugar.

Epilepsy EEG signals classification based on sparse principal component logistic regression model

Epilepsy is a chronic brain disease caused by excessive discharge of brain neurons. Long-term recurrent seizures bring a lot of trouble to patients and their families. Prediction of different stages of epilepsy is of great significance. We extract pearson correlation coefficients (PCC) between channels in different frequency bands as features of EEG signals for epilepsy stages prediction. However, the features are of large feature dimension and serious multi-collinearity. To eliminate these adverse influence, the combination of traditional dimension reduction method principal component analysis (PCA) and logistic regression method with regularization term is proposed to avoid over-fitting and achieve the feature sparsity. The experiments are conducted on the widely used CHB-MIT dataset using different regularization terms L 1 and L 2 , respectively. The proposed method identifies various stages of epilepsy quickly and efficiently, and it presents the best average accuracy of 94.86%, average precision of 96.71%, average recall of 93.48%, average kappa value of 0.90 and average Matthews correlation coefficient (MCC) value of 0.90 for all patients.

Extraction of Coastal Levees Using U-Net Model with Visible and Topographic Images Observed by High-Resolution Satellite Sensors.

Coastal levees play a role in protecting coastal areas from storm surges and high waves, and they provide important input information for inundation damage simulations. However, coastal levee data with uniformity and sufficient accuracy for inundation simulations are not always well developed. Against this background, this study proposed a method to extract coastal levees by inputting high spatial resolution optical satellite image products (RGB images, digital surface models (DSMs), and slope images that can be generated from DSM images), which have high data availability at the locations and times required for simulation, into a deep learning model. The model is based on U-Net, and post-processing for noise removal was introduced to further improve its accuracy. We also proposed a method to calculate levee height using a local maximum filter by giving DSM values to the extracted levee pixels. The validation was conducted in the coastal area of Ibaraki Prefecture in Japan as a test area. The levee mask images for training were manually created by combining these data with satellite images and Google Street View, because the levee GIS data created by the Ibaraki Prefectural Government were incomplete in some parts. First, the deep learning models were compared and evaluated, and it was shown that U-Net was more accurate than Pix2Pix and BBS-Net in identifying levees. Next, three cases of input images were evaluated: (Case 1) RGB image only, (Case 2) RGB and DSM images, and (Case 3) RGB, DSM, and slope images. Case 3 was found to be the most accurate, with an average Matthews correlation coefficient of 0.674. The effectiveness of noise removal post-processing was also demonstrated. In addition, an example of the calculation of levee heights was presented and evaluated for validity. In conclusion, this method was shown to be effective in extracting coastal levees. The evaluation of generalizability and use in actual inundation simulations are future tasks.

Development of a standardized methodology for transfer learning with QSAR models: a purely data-driven approach for source task selection

ABSTRACT Transfer learning is a machine learning technique that works well with chemical endpoints, with several papers confirming its efficiency. Although effective, because the choice of source/assistant tasks is non-trivial, the application of this technique is severely limited by the domain knowledge of the modeller. Considering this limitation, we developed a purely data-driven approach for source task selection that abstracts the need for domain knowledge. To achieve this, we created a supervised learning setting in which transfer outcome (positive/negative) is the variable to be predicted, and a set of six transferability metrics, calculated based on information from target and source datasets, are the features for prediction. We used the ChEMBL database to generate 100,000 transfers using random pairing, and with these transfers, we trained and evaluated our transferability prediction model (TP-Model). Our TP-Model achieved a 135-fold increase in precision while achieving a sensitivity of 92%, demonstrating a clear superiority against random search. In addition, we observed that transfer learning could provide considerable performance increases when applicable, with an average Matthews Correlation Coefficient (MCC) increase of 0.19 when using a single source and an average MCC increase of 0.44 when using multiple sources.

WE-BA: Covid-19 detection by Wavelet Entropy and Bat Algorithm

Covid-19 is a kind of fast-spreading pneumonia and has dramatically impacted human life and the economy. As early diagnosis is the most effective method to treat patients and block virus transmission, an accurate, automatic, and effective diagnosis method is needed. Our research proposes a machine learning model (WE-BA) using wavelet entropy for feature extraction to reduce the excessive features, one-layer FNNs for classification, 10-fold cross-validation (CV) to reuse the data for the relatively small dataset, and bat algorithm (BA) as a training algorithm. The experiment eventually achieved excellent performance with an average sensitivity of 75.27% ± 3.25%, an average specificity of 75.88% ± 1.89%, an average precision of 75.75% ± 1.06%, an average accuracy of 75.57% ± 1.21%, an average F1 score of 75.47% ± 1.64%, an average Matthews correlation coefficient of 51.20% ± 2.42%, and an average Fowlkes–Mallows index of 75.49% ± 1.64%. The experiments showed that the proposed WE-BA method yielded superior performance to the state-of-the-art methods. The results also proved the potential of the proposed method for the CT image classification task of Covid-19 on a small dataset.

CGUFS: A clustering-guided unsupervised feature selection algorithm for gene expression data

(Aim) Gene expression data is typically high dimensional with a limited number of samples and contain many features that are unrelated to the disease of interest. Existing unsupervised feature selection algorithms primarily focus on the significance of features in maintaining the data structure while not taking into account the redundancy among features. Determining the appropriate number of significant features is another challenge. (Method) In this paper, we propose a clustering-guided unsupervised feature selection (CGUFS) algorithm for gene expression data that addresses these problems. Our proposed algorithm introduces three improvements over existing algorithms. For the problem that existing clustering algorithms require artificially specifying the number of clusters, we propose an adaptive k-value strategy to assign appropriate pseudo-labels to each sample by iteratively updating a change function. For the problem that existing algorithms fail to consider the redundancy among features, we propose a feature grouping strategy to group highly redundant features. For the problem that the existing algorithms cannot filter the redundant features, we propose an adaptive filtering strategy to determine the feature combinations to be retained by calculating the potentially effective features and potentially redundant features of each feature group. (Result) Experimental results show that the average accuracy (ACC) and matthews correlation coefficient (MCC) indexes of the C4.5 classifier on the optimal features selected by the CGUFS algorithm reach 74.37% and 63.84%, respectively, significantly superior to the existing algorithms. (Conclusion) Similarly, the average ACC and MCC indexes of the Adaboost classifier on the optimal features selected by the CGUFS algorithm are significantly superior to the existing algorithms. In addition, statistical experiment results show significant differences between the CGUFS algorithm and the existing algorithms.

Improving DNA 6mA Site Prediction via Integrating Bidirectional Long Short-Term Memory, Convolutional Neural Network, and Self-Attention Mechanism.

Identifying DNA N6-methyladenine (6mA) sites is significantly important to understanding the function of DNA. Many deep learning-based methods have been developed to improve the performance of 6mA site prediction. In this study, to further improve the performance of 6mA site prediction, we propose a new meta method, called Co6mA, to integrate bidirectional long short-term memory (BiLSTM), convolutional neural networks (CNNs), and self-attention mechanisms (SAM) via assembling two different deep learning-based models. The first model developed in this study is called CBi6mA, which is composed of CNN, BiLSTM, and fully connected modules. The second model is borrowed from LA6mA, which is an existing 6mA prediction method based on BiLSTM and SAM modules. Experimental results on two independent testing sets of different model organisms, i.e., Arabidopsis thaliana and Drosophila melanogaster, demonstrate that Co6mA can achieve an average accuracy of 91.8%, covering 89% of all 6mA samples while achieving an average Matthews correlation coefficient value (0.839), which is higher than the second-best method DeepM6A.

Prediction of plasma trough concentration of voriconazole in adult patients using machine learning

ObjectivePlasma trough concentration of voriconazole (VCZ) was associated with its toxicity and efficacy. However, the nonlinear pharmacokinetic characteristics of VCZ make it difficult to determine the relationship between clinical characteristics and its concentration. We intended to present a machine learning (ML)-based method to predict toxic plasma trough concentration of VCZ (>5 μg/mL). MethodsA single center retrospective study was conducted. Three ML algorithms were used to estimate the concentration in adult patients, including random forest (RF), gradient boosting (GB), and extreme gradient boosting (XGBoost). The importance of variables was recognized by the SHapley Additive exPlanations (SHAP) method. In addition, an external validation set was used to validate the robustness of models. ResultsA total of 1318 VCZ plasma concentration were included, with 33 variables enrolled in the model. Nine classification models were developed using the RF, GB, and XGBoost algorithms. Most models performed well for both the training set and test set, with an average balanced accuracy (BA) of 0.704 and an average accuracy (ACC) of 0.788. In addition, the average Matthews correlation coefficient value reached 0.484, which indicated the predicted values are meaningful. Based on the average BA and ACC values, the predictive ability of the models can be ranked from best to worst as follows: younger adult models > mixed models > elderly models, and XGBoost models > GBT models > RF models. The SHAP results showed that the top five influencing factors in younger adult patients (<60 years) were albumin, total bile acid (TBA), platelets count, age, and inflammation, while the top five influencing factors in elderly patients were albumin, TBA, aspartate aminotransferase, creatinine, and alanine aminotransferase. Furthermore, the prediction of external validation set for VCZ concentrations verified the high reliability of the models, for the ACC value of 0.822 by the best model. ConclusionsThe ML models can be reliable tools for predicting toxic concentration exposure of VCZ. The SHAP results may provide useful guidelines for dosage adjustment of VCZ.

Improving protein-protein interaction site prediction using deep residual neural network

Accurate identification of protein-protein interaction (PPI) sites is significantly important for understanding the mechanism of life and developing new drugs. However, it is expensive and time-consuming to identify PPI sites using wet-lab experiments. Developing computational methods is a new road to identify PPI sites, which can accelerate the procedure of PPI-related research. In this study, we propose a novel deep learning-based method (called D-PPIsite) to improve the accuracy of sequence-based PPI site prediction. In D-PPIsite, four discriminative sequence-driven features, i.e., position specific scoring matrix, relative solvent accessibility, position information and physical properties, are employed to feed into a well-designed deep learning module, consisting of convolutional, squeeze and excitation, and fully connected layers, to learn a prediction model. To reduce the risk of a single prediction model getting stuck in local optima, multiple prediction models with different initialization parameters are selected and integrated into one final model using the mean ensemble strategy. Experimental results on five independent testing data sets demonstrate that the proposed D-PPIsite can achieve an average accuracy of 80.2% and precision of 36.9%, covering 53.5% of all PPI sites while achieving the average Matthews correlation coefficient value (0.330) that is significantly higher than most of existing state-of-the-art prediction methods. We implement a new standalone-version predictor for predicting PPI sites, which is freely available at https://github.com/MingDongup/D-PPIsite for academic use.

Automating the Standardized Cosmesis and Health Nasal Outcomes Survey Classification with Convolutional Neural Networks.

Importance: Currently, the aesthetic appearance and structure of the nose in a rhinoplasty patient is evaluated by a surgeon, without automation. Objective: To compare the assessment of convolutional neural networks (CNNs) (machine learning) and a rhinoplasty surgeon's impression of the nose before rhinoplasty. Methods: Preoperative nasal images were scored using a modified standardized cosmesis and health nasal outcomes survey (SCHNOS) questionnaire. Artificial intelligence (AI) models based on CNNs were developed and trained to classify patient nasal aesthetics into one of five categories, representing even intervals on the SCHNOS scoring scale. The models' performances were benchmarked against expert surgeon evaluation. Results: Two hundred thirty-five preoperative patient images were included in the study. The best-performing AI model achieved 61% accuracy and 0.449 average Matthews Correlation Coefficient on new patients. Conclusions: This pilot study suggests a proof-of-concept for AI to allow an automated patient assessment tool trained on preoperative patient images with a potential utility for counseling rhinoplasty patients.

Object-Based Burned Area Mapping with Extreme Gradient Boosting Using Sentinel-2 Imagery

The Sentinel-2 satellites are providing an unparalleled wealth of high-resolution remotely sensed information with a short revisit cycle, which is ideal for mapping burned areas both accurately and timely. This paper proposes an automated methodology for mapping burn scars using pairs of Sentinel-2 imagery, exploiting the state-of-the-art eXtreme Gradient Boosting (XGB) machine learning framework. A large database of 64 reference wildfire perimeters in Greece from 2016 to 2019 is used to train the classifier. An empirical methodology for appropriately sampling the training patterns from this database is formulated, which guarantees the effectiveness of the approach and its computational efficiency. A difference (pre-fire minus post-fire) spectral index is used for this purpose, upon which we appropriately identify the clear and fuzzy value ranges. To reduce the data volume, a super-pixel segmentation of the images is also employed, implemented via the QuickShift algorithm. The cross-validation results showcase the effectiveness of the proposed algorithm, with the average commission and omission errors being 9% and 2%, respectively, and the average Matthews correlation coefficient (MCC) equal to 0.93.

A deep learning approach for orphan gene identification in moso bamboo (Phyllostachys edulis) based on the CNN + Transformer model.

BackgroundOrphan gene play an important role in the environmental stresses of many species and their identification is a critical step to understand biological functions. Moso bamboo has high ecological, economic and cultural value. Studies have shown that the growth of moso bamboo is influenced by various stresses. Several traditional methods are time-consuming and inefficient. Hence, the development of efficient and high-accuracy computational methods for predicting orphan genes is of great significance.ResultsIn this paper, we propose a novel deep learning model (CNN + Transformer) for identifying orphan genes in moso bamboo. It uses a convolutional neural network in combination with a transformer neural network to capture k-mer amino acids and features between k-mer amino acids in protein sequences. The experimental results show that the average balance accuracy value of CNN + Transformer on moso bamboo dataset can reach 0.875, and the average Matthews Correlation Coefficient (MCC) value can reach 0.471. For the same testing set, the Balance Accuracy (BA), Geometric Mean (GM), Bookmaker Informedness (BM), and MCC values of the recurrent neural network, long short-term memory, gated recurrent unit, and transformer models are all lower than those of CNN + Transformer, which indicated that the model has the extensive ability for OG identification in moso bamboo.ConclusionsCNN + Transformer model is feasible and obtains the credible predictive results. It may also provide valuable references for other related research. As our knowledge, this is the first model to adopt the deep learning techniques for identifying orphan genes in plants.Supplementary InformationThe online version contains supplementary material available at 10.1186/s12859-022-04702-1.

Phosphate binding sites prediction in phosphorylation-dependent protein-protein interactions.

Phosphate binding plays an important role in modulating protein-protein interactions, which are ubiquitous in various biological processes. Accurate prediction of phosphate binding sites is an important but challenging task. Small size and diversity of phosphate binding sites lead to a substantial challenge for developing accurate prediction methods. Here, we present the phosphate binding site predictor (PBSP), a novel and accurate approach to identifying phosphate binding sites from protein structures. PBSP combines an energy-based ligand-binding sites identification method with reverse focused docking using a phosphate probe. We show that PBSP outperforms not only general ligand binding sites predictors but also other existing phospholigand-specific binding sites predictors. It achieves ∼95% success rate for top 10 predicted sites with an average Matthews correlation coefficient value of 0.84 for successful predictions. PBSP can accurately predict phosphate binding modes, with average position error of 1.4 and 2.4 Å in bound and unbound datasets, respectively. Lastly, visual inspection of the predictions is conducted. Reasons for failed predictions are further analyzed and possible ways to improve the performance are provided. These results demonstrate a novel and accurate approach to phosphate binding sites identification in protein structures. The software and benchmark datasets are freely available at http://web.pkusz.edu.cn/wu/PBSP/. Supplementary data are available at Bioinformatics online.

DeepnsSNPs: Accurate prediction of non-synonymous single-nucleotide polymorphisms by combining multi-scale convolutional neural network and residue environment information

Non-synonymous single-nucleotide polymorphisms (nsSNPs) is a typical kind of genetic variant, and more than 6000 diseases have been detected to be caused by nsSNPs. Accordingly, the accurate prediction of nsSNPs is of great importance for a better understanding of their functional mechanisms and disease treatment. Till now, many computational studies have been developed to identify disease-causing nsSNPs from the neutral ones; however, there is still some gap existing for further improvement in terms of overall prediction performance. In this work, we proposed a novel deep learning model, called multi-scale convolutional neural network (MSCNN). It utilized multi-scale convolution with different kernel sizes for feature processing, which can collect more effective characteristics than using a single convolution kernel size. Moreover, we applied three types of nominal structural features for further improving the nsSNPs prediction performance. Notably, the nsSNPs sequence and structural features were extracted based on the “residue environment” method we proposed, which has proved to be effective for protein nsSNPs prediction in our previous research. Based on the proposed MSCNN model and the extracted informative feature matrix, we implemented a new nsSNPs predictor, named DeepnsSNPs. The DeepnsSNPs was tested on three nsSNPs datasets collected from the PredictSNP1 website and achieved an average Matthews correlation coefficient of 0.507, which is 18.28% higher than the individual classifiers and 11.37% higher than the consensus classifier on average. Detailed dataset analyses have demonstrated that the DeepnsSNPs would be useful in the nsSNPs prediction. We provide the source python codes and benchmark datasets at https://github.com/sera616/DeepnsSNPs.git for academic use.

A novel fusion strategy for locomotion activity recognition based on multimodal signals

Detecting locomotion activities and gait phases are critical for recognizing movement intent and assisting humans in activities of daily living. This paper introduces a novel fusion strategy that analyzes multimodal signals (electromyograms, angular rates, and footswitches) to identify four lower limb locomotion activities (walking, running, stair ascent, and descent), four gait phases (push-up, stance, step-up, and swing) and detect gait events (heel strike and toe-off). This strategy uses different networks (convolutional neural network and Bi-directional Long Short-Term Memory) and a feature fusion method (Canonical Correlation Analysis) to perform the classification tasks and gait event detection, which transcends the limitations of conventional methods that only analyze a single signal. Then, an adaptive online algorithm is proposed, which helps the model to automatically fine-tune its parameters and gradually approach the convergence state. The offline experiment results on 20 healthy subjects show an average accuracy of 99.33 % and an average Matthews correlation coefficient of 99.56. Besides, with the help of the adaptive online algorithm, lower errors are shown in gait event detection, including 18.43±7.38 ms for heel strike and 22.25±7.06 ms for toe-off. Consequently, our proposed model can provide more accurate quantitative results than mainstream methods, giving it a competitive advantage for future applications.

Deep learning with convolutional neural network in the assessment of breast cancer molecular subtypes based on US images: a multicenter retrospective study.

To evaluate the prediction performance of deep convolutional neural network (DCNN) based on ultrasound (US) images for the assessment of breast cancer molecular subtypes. A dataset of 4828 US images from 1275 patients with primary breast cancer were used as the training samples. DCNN models were constructed primarily to predict the four St. Gallen molecular subtypes and secondarily to identify luminal disease from non-luminal disease based on the ground truth from immunohistochemical of whole tumor surgical specimen. US images from two other institutions were retained as independent test sets to validate the system. The models' performance was analyzed using per-class accuracy, positive predictive value (PPV), and Matthews correlation coefficient (MCC). The model achieved good performance in identifying the four breast cancer molecular subtypes in the two test sets, with accuracy ranging from 80.07% (95% CI, 76.49-83.23%) to 97.02% (95% CI, 95.22-98.16%) and 87.94% (95% CI, 85.08-90.31%) to 98.83% (95% CI, 97.60-99.43) for the two test cohorts for each sub-category, respectively. In terms of 4-class weighted average MCC, the model achieved 0.59 for test cohort A and 0.79 for test cohort B. Specifically, the DCNN also yielded good diagnostic performance in discriminating luminal disease from non-luminal disease, with a PPV of 93.29% (95% CI, 90.63-95.23%) and 88.21% (95% CI, 85.12-90.73%) for the two test cohorts, respectively. Using pretreatment US images of the breast cancer, deep learning model enables the assessment of molecular subtypes with high diagnostic accuracy. Clinical trial number: ChiCTR1900027676 KEY POINTS: • Deep convolutional neural network (DCNN) helps clinicians assess tumor features with accuracy. • Multicenter retrospective study shows that DCNN derived from pretreatment ultrasound imagine improves the prediction of breast cancer molecular subtypes. • Management of patients becomes more precise based on the DCNN model.

Deep Belief Network for Predicting the Predisposition to Lung Cancer in TP53 Gene

Lung cancer, similar to other cancer types, results from genetic changes. However, it is considered as more threatening due to the spread of the smoking habit, a major risk factor of the disease. Scientists have been collecting and analyzing the biological data for a long time, in attempts to find methods to predict cancer before it occurs. Analysis of these data requires the use of artificial intelligence algorithms and neural network approaches. In this paper, one of the deep neural networks was used, that is the enhancer Deep Belief Network (DBN), which is constructed from two Restricted Boltzmann Machines (RBM). The visible nodes for the first RBM are 13 nodes and 8 nodes in each hidden layer for the two RBMs. The enhancer DBN was trained by Back Propagation Neural Network (BPNN), where the data sets were divided into 6 folds, each is split into three partitions representing the training, validation, and testing. It is worthy to note that the proposed enhancer DBN predicted lung cancer in an acceptable manner, with an average F-measure value of  0. 96 and an average Matthews Correlation Coefficient (MCC) value of 0. 47 for 6 folds.

Object-Based Image Analysis Applied to Low Altitude Aerial Imagery for Potato Plant Trait Retrieval and Pathogen Detection.

There is a growing demand in both food quality and quantity, but as of now, one-third of all food produced for human consumption is lost due to pests and other pathogens accounting for roughly 40% of pre-harvest loss in potatoes. Pathogens in potato plants, like the Erwinia bacteria and the PVYNTN virus for example, exhibit symptoms of varying severity that are not easily captured by pixel-based classes (as these ignore shape, texture, and context in general). The aim of this research is to develop an object-based image analysis (OBIA) method for trait retrieval of individual potato plants that maximizes information output from Unmanned Aerial Vehicle (UAV) RGB very high resolution (VHR) imagery and its derivatives, to be used for disease detection of the Solanum tuberosum. The approach proposed can be split in two steps: (1) object-based mapping of potato plants using an optimized implementation of large scale mean-shift segmentation (LSMSS), and (2) classification of disease using a random forest (RF) model for a set of morphological traits computed from their associative objects. The approach was proven viable as the associative RF model detected presence of Erwinia and PVY pathogens with a maximum F1 score of 0.75 and an average Matthews Correlation Coefficient (MCC) score of 0.47. It also shows that low-altitude imagery acquired with a commercial UAV is a viable off-the-shelf tool for precision farming, and potato pathogen detection.

Prediction of novel mouse\xa0TLR9 agonists using a random forest approach

BackgroundToll-like receptor 9 is a key innate immune receptor involved in detecting infectious diseases and cancer. TLR9 activates the innate immune system following the recognition of single-stranded DNA oligonucleotides (ODN) containing unmethylated cytosine-guanine (CpG) motifs. Due to the considerable number of rotatable bonds in ODNs, high-throughput in silico screening for potential TLR9 activity via traditional structure-based virtual screening approaches of CpG ODNs is challenging. In the current study, we present a machine learning based method for predicting novel mouse TLR9 (mTLR9) agonists based on features including count and position of motifs, the distance between the motifs and graphically derived features such as the radius of gyration and moment of Inertia. We employed an in-house experimentally validated dataset of 396 single-stranded synthetic ODNs, to compare the results of five machine learning algorithms. Since the dataset was highly imbalanced, we used an ensemble learning approach based on repeated random down-sampling.ResultsUsing in-house experimental TLR9 activity data we found that random forest algorithm outperformed other algorithms for our dataset for TLR9 activity prediction. Therefore, we developed a cross-validated ensemble classifier of 20 random forest models. The average Matthews correlation coefficient and balanced accuracy of our ensemble classifier in test samples was 0.61 and 80.0%, respectively, with the maximum balanced accuracy and Matthews correlation coefficient of 87.0% and 0.75, respectively. We confirmed common sequence motifs including ‘CC’, ‘GG’,‘AG’, ‘CCCG’ and ‘CGGC’ were overrepresented in mTLR9 agonists. Predictions on 6000 randomly generated ODNs were ranked and the top 100 ODNs were synthesized and experimentally tested for activity in a mTLR9 reporter cell assay, with 91 of the 100 selected ODNs showing high activity, confirming the accuracy of the model in predicting mTLR9 activity.ConclusionWe combined repeated random down-sampling with random forest to overcome the class imbalance problem and achieved promising results. Overall, we showed that the random forest algorithm outperformed other machine learning algorithms including support vector machines, shrinkage discriminant analysis, gradient boosting machine and neural networks. Due to its predictive performance and simplicity, the random forest technique is a useful method for prediction of mTLR9 ODN agonists.