Machine Learning-based Scoring Research Articles

Reassessment of marker genes in human induced pluripotent stem cells for enhanced quality control

Human induced pluripotent stem cells (iPSCs) have great potential in research, but pluripotency testing faces challenges due to non-standardized methods and ambiguous markers. Here, we use long-read nanopore transcriptome sequencing to discover 172 genes linked to cell states not covered by current guidelines. We validate 12 genes by qPCR as unique markers for specific cell fates: pluripotency (CNMD, NANOG, SPP1), endoderm (CER1, EOMES, GATA6), mesoderm (APLNR, HAND1, HOXB7), and ectoderm (HES5, PAMR1, PAX6). Using these genes, we develop a machine learning-based scoring system, “hiPSCore”, trained on 15 iPSC lines and validated on 10 more. hiPSCore accurately classifies pluripotent and differentiated cells and predicts their potential to become specialized 2D cells and 3D organoids. Our re-evaluation of cell fate marker genes identifies key targets for future studies on cell fate assessment. hiPSCore improves iPSC testing by reducing time, subjectivity, and resource use, thus enhancing iPSC quality for scientific and medical applications.

NeoMS: Mass Spectrometry-based Method for Uncovering Mutated MHC-I Neoantigens.

Major Histocompatibility Complex (MHC) molecules play a critical role in the immune system by presenting peptides on the cell surface for recognition by T-cells. Tumor cells often produce MHC peptides with amino acid mutations, known as neoantigens, which evade T-cell recognition, leading to rapid tumor growth. In immunotherapies such as TCR-T and CAR-T, identifying these mutated MHC peptide sequences is crucial. Current mass spectrometry-based peptide identification methods primarily rely on database searching, which fails to detect mutated peptides not present in human databases. In this paper, we propose a novel workflow called NeoMS, designed to efficiently identify both non-mutated and mutated MHC-I peptides from mass spectrometry data. NeoMS utilizes a tagging algorithm to generate an expanded sequence database that includes potential mutated proteins for each sample. Furthermore, it employs a machine learning-based scoring function for each peptide-spectrum match (PSM) to maximize search sensitivity. Finally, a rigorous target-decoy approach is implemented to control the false discovery rates (FDR) of the peptides with and without mutations separately. Experimental results for regular peptides demonstrate that NeoMS outperforms four benchmark methods. For mutated peptides, NeoMS successfully identifies hundreds of high-quality mutated peptides in a melanoma-associated sample, with their validity confirmed by further studies.

Promoter mutation-independent TERT expression is related to the immune-enriched milieu in papillary thyroid cancer.

Telomerase reverse transcriptase promoter mutation (pTERT MT) promotes human carcinogenesis via aberrant expression of telomerase reverse transcriptase (TERT). However, the tumorigenic impact of TERT expression independent of pTERT MT remains unclear despite numerous mechanisms of TERT being suggested. To tackle this issue, we employed comprehensive bioinformatics to assess biological variations noticed among different TERT expression mechanisms. Papillary thyroid cancer (PTC) with pTERT MT (pTERT MT PTC) presented aggressive clinical behavior and exhibited biological profiles associated with cellular immortality and genomic instability. PTC with TERT expression but without pTERT MT (TERT (+) PTC), also exhibited poor clinicopathological characteristics and was enriched with immune responses. In accordance, c-MYC/E2F and nuclear factor kappa B (NFκB) were dominant transcription factors in pTERT MT PTC and TERT (+) PTC, respectively. Notably, we revealed TERT hypermethylated oncological region (THOR) as a potential TERT expressing mechanism in TERT (+) PTC patients. Furthermore, three unique subtypes of papillary thyroid cancer were deciphered using a combination of machine learning-based scoring systems. Our proposed scoring system was clinically significant, especially in microcarcinoma, predicting survival outcomes and inferring therapeutic responses to radioactive iodine therapy. Finally, our analysis was expanded to endocrine-related cancers, unveiling various regulatory mechanisms of TERT with poor clinical outcomes and biological behaviors.

An ensemble-based approach to estimate confidence of predicted protein-ligand binding affinity values.

When designing a machine learning-based scoring function, we access a limited number of protein-ligand complexes with experimentally determined binding affinity values, representing only a fraction of all possible protein-ligand complexes. Consequently, it is crucial to report a measure of confidence and quantify the uncertainty in the model's predictions during test time. Here, we adopt the conformal prediction technique to evaluate the confidence of a prediction for each member of the core set of the CASF 2016 benchmark. The conformal prediction technique requires a diverse ensemble of predictors for uncertainty estimation. To this end, we introduce ENS-Score as an ensemble predictor, which includes 30 models with different protein-ligand representation approaches and achieves Pearson's correlation of 0.842 on the core set of the CASF 2016 benchmark. Also, we comprehensively investigate the residual error of each data point to assess the normality behavior of the distribution of the residual errors and their correlation to the structural features of the ligands, such as hydrophobic interactions and halogen bonding. In the end, we provide a local host web application to facilitate the usage of ENS-Score. All codes to repeat results are provided at https://github.com/miladrayka/ENS_Score.

CADD v1.7: using protein language models, regulatory CNNs and other nucleotide-level scores to improve genome-wide variant predictions.

Machine Learning-based scoring and classification of genetic variants aids the assessment of clinical findings and is employed to prioritize variants in diverse genetic studies and analyses. Combined Annotation-Dependent Depletion (CADD) is one of the first methods for the genome-wide prioritization of variants across different molecular functions and has been continuously developed and improved since its original publication. Here, we present our most recent release, CADD v1.7. We explored and integrated new annotation features, among them state-of-the-art protein language model scores (Meta ESM-1v), regulatory variant effect predictions (from sequence-based convolutional neural networks) and sequence conservation scores (Zoonomia). We evaluated the new version on data sets derived from ClinVar, ExAC/gnomAD and 1000 Genomes variants. For coding effects, we tested CADD on 31 Deep Mutational Scanning (DMS) data sets from ProteinGym and, for regulatory effect prediction, we used saturation mutagenesis reporter assay data of promoter and enhancer sequences. The inclusion of new features further improved the overall performance of CADD. As with previous releases, all data sets, genome-wide CADD v1.7 scores, scripts for on-site scoring and an easy-to-use webserver are readily provided via https://cadd.bihealth.org/ or https://cadd.gs.washington.edu/ to the community.

A machine learning-based scoring system and ten factors associated with hip fracture occurrence in the elderly.

Hip fractures are fragility fractures frequently seen in persons over 80-years-old. Although various factors, including decreased bone mineral density and a history of falls, are reported as hip fracture risks, few large-scale studies have confirmed their relevance to individuals older than 80, and tools to assess contributions of various risks to fracture development and the degree of risk are lacking. We recruited 1395 fresh hip fracture patients and 1075 controls without hip fractures and comprehensively evaluated various reported risk factors and their association with hip fracture development. We initially constructed a predictive model using Extreme Gradient Boosting (XGBoost), a machine learning algorithm, incorporating all 40 variables and evaluated the model's performance using the area under the receiver operating characteristic curve (AUC), yielding a value of 0.87. We also employed SHapley Additive exPlanation (SHAP) values to evaluate each feature importance and ranked the top 20. We then used a stepwise selection method to determine key factors sequentially until the AUC reached a plateau nearly equal to that of all variables and identified the top 10 sufficient to evaluate hip fracture risk. For each, we determined the cutoff value for hip fracture occurrence and calculated scores of each variable based on the respective feature importance. Individual scores were: serum 25(OH)D levels (<10ng/ml, score 7), femoral neck T-score (<-3, score 5), Barthel index score (<100, score 3), maximal handgrip strength (<18kg, score 3), GLFS-25 score (≥24, score 2), number of falls in previous 12months (≥3, score 2), serum IGF-1 levels (<50ng/ml, score 2), cups of tea/day (≥5, score -2), use of anti-osteoporosis drugs (yes, score -2), and BMI (<18.5kg/m2, score 1). Using these scores, we performed receiver operating characteristic (ROC) analysis and the resultant optimal cutoff value was 7, with a specificity of 0.78, sensitivity of 0.75, and AUC of 0.85. These ten factors and the scoring system may represent tools useful to predict hip fracture.

A Novel Machine Learning-Based Scoring System for Estimating the Risk of Chronic Kidney Disease in Community-Dwelling Elderly

&lt;p&gt;目的：本研究致力於運用新型機器學習技術，開發出能準確預測老年族群慢性腎臟病(chronic kidney disease, CKD)風險、同時擁有良好風險校準能力的CKD風險分級模型，以期能於臨床情境快速辨識出需加強照護管理的CKD高風險老年族群。&lt;/p&gt; &lt;p&gt;方法：本研究採用回溯性研究方法，分析台灣北部某教學醫院從2014年至2018年，3,469名受檢者的臨床相關數據，篩選出與CKD具較高關聯性的17個候選變量，運用Risk-Calibrated Supersparse Linear Integer Model（簡稱RiskSLIM）的機器學習技術，構建新型CKD風險分級模型。&lt;/p&gt; &lt;p&gt;結果：在3,469名受試者中，共有463名(13.3%)受試者符合Kidney Disease Outcomes and Quality Initiative（簡稱KDOQI）所定義的CKD標準。經實驗發現，在區分CKD與非 CKD受檢者的任務中，具有五項變量的RiskSLIM候選模型具有最佳的表現，命名為CHUME風險分數(Creatinine-Hypertension-bUn-Male-Exercise score, CHUME score)。與傳統模型相比，RiskSLIM模型具有更佳的風險校準能力（RiskSLIM模型的平均5-CV CAL為3.6% [95% CI 3.1%-4.1%]，對比Penalized logistic regression模型為4.7% [95% CI 4.3%-5.0%]）。若將CHUME風險分數的診斷閾值設定在總分大於等於1分時判斷為陽性，會有最佳的分類預測表現，AUC為0.912 [95% CI 0.896-0.927]；敏感度為83.8% [95% CI 80.4%-87.1%]；特異度為87.7% [95% CI 86.5%-88.8%]。&lt;/p&gt; &lt;p&gt;結論：CHUME風險分數是使用RiskSLIM機器學習技術構建出的CKD風險分級模型，易於使用，同時具有高敏感度及高特異度，能夠快速整合入現有的CKD臨床診療流程中，是一個適合用於社區老年族群的CKD 臨床篩檢工具。&lt;/p&gt; &lt;p&gt;&amp;nbsp;&lt;/p&gt;&lt;p&gt;Purpose: In response to the increasing need of CKD (chronic kidney disease) management, we developed a novel machine learning-based screening system to better assess CKD risk among community-dwelling older adults. &lt;/p&gt; &lt;p&gt; Methods: Based on a retrospective research design, our study analyzed the clinical data of 3,469 senior citizens receiving health checkups during the five year from 2014 to 2018 to identify 17 candidate variables for inclusion in the CKD risk screening model, using the machine learning technique named Risk-Calibrated Supersparse Linear Integer Model (RiskSLIM). &lt;/p&gt; &lt;p&gt; Results: Of the total 3,469 participants, 463 (13.3%) met the KDOQI-CKD criteria. A 5-item RiskSLIM model was found to be optimal in measuring the creatinine-hypertension-bUn-male-exercise (CHUME) score for distinguishing CKD from non-CKD cases. In contrast to the baseline Penalized logistic regression (PLR) model, the RiskSLIM model was simpler and superior in terms of accuracy in risk calibration (mean 5-CV CAL of 3.6% [95% CI 3.1%-4.1%] vs 4.7% [95% CI 4.3%-5.0%] for PLR). Operating characteristics for CHUME score with 95% CIs developed from bootstrapping from all participants were remarkable at the diagnostic threshold (score of 1-or-higher) with an estimated AUC of 0.912 [95%-CI 0.896-0.927], sensitivity of 83.8% [95%-CI 80.4%-87.1%], and specificity of 87.7% [95%-CI 86.5%-88.8%]. &lt;/p&gt; &lt;p&gt; Conclusion: The RiskLIM-based CKD screening system is easy to use, and the CHUME score measured is marked with high sensitivity and specificity. The simplicity, as well as the efficacy, of the screening system is conducive to easy integration into clinical workflow to help assess CKD risk. We accordingly conclude that it is a suitable CKD screening tool for community-dwelling elderly.&lt;/p&gt; &lt;p&gt;&amp;nbsp;&lt;/p&gt;

ML-PLIC: a web platform for characterizing protein-ligand interactions and developing machine learning-based scoring functions.

Cracking the entangling code of protein-ligand interaction (PLI) is of great importance to structure-based drug design and discovery. Different physical and biochemical representations can be used to describe PLI such as energy terms and interaction fingerprints, which can be analyzed by machine learning (ML) algorithms to create ML-based scoring functions (MLSFs). Here, we propose the ML-based PLI capturer (ML-PLIC), a web platform that automatically characterizes PLI and generates MLSFs to identify the potential binders of a specific protein target through virtual screening (VS). ML-PLIC comprises five modules, including Docking for ligand docking, Descriptors for PLI generation, Modeling for MLSF training, Screening for VS and Pipeline for the integration of the aforementioned functions. We validated the MLSFs constructed by ML-PLIC in three benchmark datasets (Directory of Useful Decoys-Enhanced, Active as Decoys and TocoDecoy), demonstrating accuracy outperforming traditional docking tools and competitive performance to the deep learning-based SF, and provided a case study of the Serine/threonine-protein kinase WEE1 in which MLSFs were developed by using the ML-based VS pipeline in ML-PLIC. Underpinning the latest version of ML-PLIC is a powerful platform that incorporates physical and biological knowledge about PLI, leveraging PLI characterization and MLSF generation into the design of structure-based VS pipeline. The ML-PLIC web platform is now freely available at http://cadd.zju.edu.cn/plic/.

Water Network-Augmented Two-State Model for Protein-Ligand Binding Affinity Prediction.

Water network rearrangement from the ligand-unbound state to the ligand-bound state is known to have significant effects on the protein-ligand binding interactions, but most of the current machine learning-based scoring functions overlook these effects. In this study, we endeavor to construct a comprehensive and realistic deep learning model by incorporating water network information into both ligand-unbound and -bound states. In particular, extended connectivity interaction features were integrated into graph representation, and graph transformer operator was employed to extract features of the ligand-unbound and -bound states. Through these efforts, we developed a water network-augmented two-state model called ECIFGraph::HM-Holo-Apo. Our new model exhibits satisfactory performance in terms of scoring, ranking, docking, screening, and reverse screening power tests on the CASF-2016 benchmark. In addition, it can achieve superior performance in large-scale docking-based virtual screening tests on the DEKOIS2.0 data set. Our study highlights that the use of a water network-augmented two-state model can be an effective strategy to bolster the robustness and applicability of machine learning-based scoring functions, particularly for targets with hydrophilic or solvent-exposed binding pockets.

TB-IECS: an accurate machine learning-based scoring function for virtual screening

Machine learning-based scoring functions (MLSFs) have shown potential for improving virtual screening capabilities over classical scoring functions (SFs). Due to the high computational cost in the process of feature generation, the numbers of descriptors used in MLSFs and the characterization of protein–ligand interactions are always limited, which may affect the overall accuracy and efficiency. Here, we propose a new SF called TB-IECS (theory-based interaction energy component score), which combines energy terms from Smina and NNScore version 2, and utilizes the eXtreme Gradient Boosting (XGBoost) algorithm for model training. In this study, the energy terms decomposed from 15 traditional SFs were firstly categorized based on their formulas and physicochemical principles, and 324 feature combinations were generated accordingly. Five best feature combinations were selected for further evaluation of the model performance in regard to the selection of feature vectors with various length, interaction types and ML algorithms. The virtual screening power of TB-IECS was assessed on the datasets of DUD-E and LIT-PCBA, as well as seven target-specific datasets from the ChemDiv database. The results showed that TB-IECS outperformed classical SFs including Glide SP and Dock, and effectively balanced the efficiency and accuracy for practical virtual screening.

A Small Step Toward Generalizability: Training a Machine Learning Scoring Function for Structure-Based Virtual Screening.

Over the past few years, many machine learning-based scoring functions for predicting the binding of small molecules to proteins have been developed. Their objective is to approximate the distribution which takes two molecules as input and outputs the energy of their interaction. Only a scoring function that accounts for the interatomic interactions involved in binding can accurately predict binding affinity on unseen molecules. However, many scoring functions make predictions based on data set biases rather than an understanding of the physics of binding. These scoring functions perform well when tested on similar targets to those in the training set but fail to generalize to dissimilar targets. To test what a machine learning-based scoring function has learned, input attribution, a technique for learning which features are important to a model when making a prediction on a particular data point, can be applied. If a model successfully learns something beyond data set biases, attribution should give insight into the important binding interactions that are taking place. We built a machine learning-based scoring function that aimed to avoid the influence of bias via thorough train and test data set filtering and show that it achieves comparable performance on the Comparative Assessment of Scoring Functions, 2016 (CASF-2016) benchmark to other leading methods. We then use the CASF-2016 test set to perform attribution and find that the bonds identified as important by PointVS, unlike those extracted from other scoring functions, have a high correlation with those found by a distance-based interaction profiler. We then show that attribution can be used to extract important binding pharmacophores from a given protein target when supplied with a number of bound structures. We use this information to perform fragment elaboration and see improvements in docking scores compared to using structural information from a traditional, data-based approach. This not only provides definitive proof that the scoring function has learned to identify some important binding interactions but also constitutes the first deep learning-based method for extracting structural information from a target for molecule design.

GB-score: Minimally designed machine learning scoring function based on distance-weighted interatomic contact features.

In recent years, thanks to advances in computer hardware and dataset availability, data-driven approaches (like machine learning) have become one of the essential parts of the drug design framework to accelerate drug discovery procedures. Constructing a new scoring function, a function that can predict the binding score for a generated protein-ligand pose during docking procedure or a crystal complex, based on machine and deep learning has become an active research area in computer-aided drug design. GB-Score is a state-of-the-art machine learning-based scoring function that utilizes distance-weighted interatomic contact features, PDBbind-v2019 general set, and Gradient Boosting Trees algorithm to the binding affinity prediction. The distance-weighted interatomic contact featurization method used the distance between different ligand and protein atom types for numerical representation of the protein-ligand complex. GB-Score attains Pearson's correlation 0.862 and RMSE 1.190 on the CASF-2016 benchmark test in the scoring power metric. GB-Score's codes are freely available on the web at https://github.com/miladrayka/GB_Score.

A generalized protein-ligand scoring framework with balanced scoring, docking, ranking and screening powers.

Applying machine learning algorithms to protein-ligand scoring functions has aroused widespread attention in recent years due to the high predictive accuracy and affordable computational cost. Nevertheless, most machine learning-based scoring functions are only applicable to a specific task, e.g., binding affinity prediction, binding pose prediction or virtual screening, suggesting that the development of a scoring function with balanced performance in all critical tasks remains a grand challenge. To this end, we propose a novel parameterization strategy by introducing an adjustable binding affinity term that represents the correlation between the predicted outcomes and experimental data into the training of mixture density network. The resulting residue-atom distance likelihood potential not only retains the superior docking and screening power over all the other state-of-the-art approaches, but also achieves a remarkable improvement in scoring and ranking performance. We emphatically explore the impacts of several key elements on prediction accuracy as well as the task preference, and demonstrate that the performance of scoring/ranking and docking/screening tasks of a certain model could be well balanced through an appropriate manner. Overall, our study highlights the potential utility of our innovative parameterization strategy as well as the resulting scoring framework in future structure-based drug design.

Machine Learning-Based Scoring System for Early Prognosis Evaluation of Patients with Coronavirus Disease 2019

Abstract Background The global spread of coronavirus disease 2019 (COVID-19) continues to threaten human health security, exerting considerable pressure on healthcare systems worldwide. While prognostic models for COVID-19 hospitalized or intensive care patients are currently available, prognostic models developed for large cohorts of thousands of individuals are still lacking. Methods Between February 4 and April 16, 2020, we enrolled 3,974 patients admitted with COVID-19 disease in the Wuhan Huo-Shen-Shan Hospital and the Maternal and Child Hospital, Hubei Province, China. (1) Screening of key prognostic factors: A univariate Cox regression analysis was performed on 2,649 patients in the training set, and factors affecting prognosis were initially screened. Subsequently, a random survival forest model was established through machine analysis to further screen for factors that are important for prognosis. Finally, multivariate Cox regression analysis was used to determine the synergy among various factors related to prognosis. (2) Establishment of a scoring system: The nomogram algorithm established a COVID-19 patient death risk assessment scoring system for the nine selected key prognostic factors, calculated the C index, drew calibration curves and drew training set patient survival curves. (3) Verification of the scoring system: The scoring system assessed 1,325 patients in the test set, splitting them into high- and low-risk groups, calculated the C-index, and drew calibration and survival curves. Results The cross-sectional study found that age, clinical classification, sex, pulmonary insufficiency, hypoproteinemia, and four other factors (underlying diseases: blood diseases, malignant tumor; complications: digestive tract bleeding, heart dysfunction) have important significance for the prognosis of the enrolled patients with COVID-19. Herein, we report the discovery of the effects of hypoproteinemia and hematological diseases on the prognosis of COVID-19. Meanwhile, the scoring system established here can effectively evaluate objective scores for the early prognoses of patients with COVID-19 and can divide them into high- and low-risk groups (using a scoring threshold of 117.77, a score below which is considered low risk). The efficacy of the system was better than that of clinical classification using the current COVID-19 guidelines (C indexes, 0.95 vs. 0.89). Conclusions Age, clinical typing, sex, pulmonary insufficiency, hypoproteinemia, and four other factors were important for COVID-19 survival. Compared with general statistical methods, this method can quickly and accurately screen out the relevant factors affecting prognosis, provide an order of importance, and establish a scoring system based on the nomogram model, which is of great clinical significance.

Machine learning-based scoring system to predict in-hospital outcomes in patients hospitalized with COVID-19

BackgroundThe evolution of patients hospitalized with coronavirus disease 2019 (COVID-19) is still hard to predict, even after several months of dealing with the pandemic. AimsTo develop and validate a score to predict outcomes in patients hospitalized with COVID-19. MethodsAll consecutive adults hospitalized for COVID-19 from February to April 2020 were included in a nationwide observational study. Primary composite outcome was transfer to an intensive care unit from an emergency department or conventional ward, or in-hospital death. A score that estimates the risk of experiencing the primary outcome was constructed from a derivation cohort using stacked LASSO (Least Absolute Shrinkage and Selection Operator), and was tested in a validation cohort. ResultsAmong 2873 patients analysed (57.9% men; 66.6±17.0 years), the primary outcome occurred in 838 (29.2%) patients: 551 (19.2%) were transferred to an intensive care unit; and 287 (10.0%) died in-hospital without transfer to an intensive care unit. Using stacked LASSO, we identified 11 variables independently associated with the primary outcome in multivariable analysis in the derivation cohort (n=2313), including demographics (sex), triage vitals (body temperature, dyspnoea, respiratory rate, fraction of inspired oxygen, blood oxygen saturation) and biological variables (pH, platelets, C-reactive protein, aspartate aminotransferase, estimated glomerular filtration rate). The Critical COVID-19 France (CCF) risk score was then developed, and displayed accurate calibration and discrimination in the derivation cohort, with C-statistics of 0.78 (95% confidence interval 0.75–0.80). The CCF risk score performed significantly better (i.e. higher C-statistics) than the usual critical care risk scores. ConclusionsThe CCF risk score was built using data collected routinely at hospital admission to predict outcomes in patients with COVID-19. This score holds promise to improve early triage of patients and allocation of healthcare resources.

Prediction of ligand binding sites using improved blind docking method with a Machine Learning-Based scoring function

The identification of the ligand binding sites (LBS) in proteins is of great significance for the elucidation of protein structure and function. Different methods have been published to predict protein–ligand binding sites efficiently. However, most of the current prediction methods only focus on the characteristics of proteins without considering the interactions between proteins and their different ligands, which often leads to frustrating results when the structure or function of the protein is complex or diverse. Therefore, an improved blind docking method with a machine learning-based scoring function is proposed in this paper for the LBS prediction. The blind docking method is used to search binding pockets and an artificial neural network is constructed to analyze binding features, which makes the proposed method possible to distinguish true binding sites from other possible pockets. Two cases of LBS prediction are presented to show the excellent performance of the proposed method. This paper aims to provide new ideas for the prediction of interactions between proteins and small molecules, which can further guide the research of structure-based drug discovery.

Improving protein-ligand docking and screening accuracies by incorporating a scoring function correction term.

Scoring functions are important components in molecular docking for structure-based drug discovery. Traditional scoring functions, generally empirical- or force field-based, are robust and have proven to be useful for identifying hits and lead optimizations. Although multiple highly accurate deep learning- or machine learning-based scoring functions have been developed, their direct applications for docking and screening are limited. We describe a novel strategy to develop a reliable protein–ligand scoring function by augmenting the traditional scoring function Vina score using a correction term (OnionNet-SFCT). The correction term is developed based on an AdaBoost random forest model, utilizing multiple layers of contacts formed between protein residues and ligand atoms. In addition to the Vina score, the model considerably enhances the AutoDock Vina prediction abilities for docking and screening tasks based on different benchmarks (such as cross-docking dataset, CASF-2016, DUD-E and DUD-AD). Furthermore, our model could be combined with multiple docking applications to increase pose selection accuracies and screening abilities, indicating its wide usage for structure-based drug discoveries. Furthermore, in a reverse practice, the combined scoring strategy successfully identified multiple known receptors of a plant hormone. To summarize, the results show that the combination of data-driven model (OnionNet-SFCT) and empirical scoring function (Vina score) is a good scoring strategy that could be useful for structure-based drug discoveries and potentially target fishing in future.

In silico binding affinity prediction for metabotropic glutamate receptors using both endpoint free energy methods and a machine learning-based scoring function.

Metabotropic glutamate receptors (mGluRs) play an important role in regulating glutamate signal pathways, which are involved in neuropathy and periphery homeostasis. mGluR4, which belongs to Group III mGluRs, is most widely distributed in the periphery among all the mGluRs. It has been proved that the regulation of this receptor is involved in diabetes, colorectal carcinoma and many other diseases. However, the application of structure-based drug design to identify small molecules to regulate the mGluR4 receptor is limited due to the absence of a resolved mGluR4 protein structure. In this work, we first built a homology model of mGluR4 based on a crystal structure of mGluR8, and then conducted hierarchical virtual screening (HVS) to identify possible active ligands for mGluR4. The HVS protocol consists of three hierarchical filters including Glide docking, molecular dynamic (MD) simulation and binding free energy calculation. We successfully prioritized active ligands of mGluR4 from a set of screening compounds using HVS. The predicted active ligands based on binding affinities can almost cover all the experiment-determined active ligands, with only one ligand missed. The correlation between the measured and predicted binding affinities is significantly improved for the MM-PB/GBSA-WSAS methods compared to the Glide docking method. More importantly, we have identified hotspots for ligand binding, and we found that SER157 and GLY158 tend to contribute to the selectivity of mGluR4 ligands, while ALA154 and ALA155 could account for the ligand selectivity to mGluR8. We also recognized other 5 key residues that are critical for ligand potency. The difference of the binding profiles between mGluR4 and mGluR8 can guide us to develop more potent and selective modulators. Moreover, we evaluated the performance of IPSF, a novel type of scoring function trained by a machine learning algorithm on residue-ligand interaction profiles, in guiding drug lead optimization. The cross-validation root-mean-square errors (RMSEs) are much smaller than those by the endpoint methods, and the correlation coefficients are comparable to the best endpoint methods for both mGluRs. Thus, machine learning-based IPSF can be applied to guide lead optimization, albeit the total number of actives/inactives are not big, a typical scenario in drug discovery projects.

FFENCODER-PL: Pair Wise Energy Descriptors for Protein-Ligand Pose Selection.

Scoring functions are the essential component in molecular docking methods. An accurate scoring function is expected to distinguish the native ligand pose from decoy poses. Our previous experience (Pei et al. J. Chem. Inf. Model. 2019, 59 (7), 3305-3315) proved that combining the random forest (RF) algorithm with knowledge-based potential functions can emphasize germane pair wise interactions and improve the performance of original knowledge-based potential functions on protein-ligand decoy detection. One of the most important potential function classes is the force field (FF) potential with one example being the Amber collection of FFs, which are widely available in the AMBER suite of simulation programs. However, for use in RF modeling studies, one needs pair wise energies that are hard to directly extract from Amber. To address this issue, FFENCODER-PL was constructed to calculate the pair wise energies based on the FF14SB and GAFF2 FFs in Amber. FFENCODER-PL was validated using 275 ligand and 21 protein-ligand structures. RF models were built by combining an RF classification algorithm with the pair wise energies calculated from FFENCODER-PL. CASF-2016 (Su et al. J. Chem. Inf. Model. 2019, 59, 895-913) was employed to test the performance of the resultant RF models, which outperformed 33 scoring functions on accuracy and native ranking tests. For the best decoy RMSD test, RF models give a best decoy with an RMSD of around 2 Å from the native pose after including the best decoy-decoy comparisons in the RF model. The relative importance of the RF algorithm and force field potentials was also tested with the results suggesting that both the RF algorithm and force field potentials are important and combining them is the only way to achieve high accuracy. Finally, FFENCODER-PL makes force field-based pair wise energies available for further development of machine learning-based scoring functions. The codes and data used in this paper can be found at https://github.com/JunPei000/Amber_protein_ligand_encoding.

A machine learning model for early detection of diabetic foot using thermogram images

Diabetes foot ulceration (DFU) and amputation are a cause of significant morbidity. The prevention of DFU may be achieved by the identification of patients at risk of DFU and the institution of preventative measures through education and offloading. Several studies have reported that thermogram images may help to detect an increase in plantar temperature prior to DFU. However, the distribution of plantar temperature may be heterogeneous, making it difficult to quantify and utilize to predict outcomes. We have compared a machine learning-based scoring technique with feature selection and optimization techniques and learning classifiers to several state-of-the-art Convolutional Neural Networks (CNNs) on foot thermogram images and propose a robust solution to identify the diabetic foot. A comparatively shallow CNN model, MobilenetV2 achieved an F1 score of ∼95% for a two-feet thermogram image-based classification and the AdaBoost Classifier used 10 features and achieved an F1 score of 97%. A comparison of the inference time for the best-performing networks confirmed that the proposed algorithm can be deployed as a smartphone application to allow the user to monitor the progression of the DFU in a home setting.