Feature selection in machine learning: A new perspective

Diabetes mellitus is a chronic disease characterized by hyperglycemia. It may cause many complications. According to the growing morbidity in recent years, in 2040, the world's diabetic patients will reach 642 million, which means that one of the ten adults in the future is suffering from diabetes. There is no doubt that this alarming figure needs great attention. With the rapid development of machine learning, machine learning has been applied to many aspects of medical health. In this study, we used decision tree, random forest and neural network to predict diabetes mellitus. The dataset is the hospital physical examination data in Luzhou, China. It contains 14 attributes. In this study, five-fold cross validation was used to examine the models. In order to verity the universal applicability of the methods, we chose some methods that have the better performance to conduct independent test experiments. We randomly selected 68994 healthy people and diabetic patients' data, respectively as training set. Due to the data unbalance, we randomly extracted 5 times data. And the result is the average of these five experiments. In this study, we used principal component analysis (PCA) and minimum redundancy maximum relevance (mRMR) to reduce the dimensionality. The results showed that prediction with random forest could reach the highest accuracy (ACC = 0.8084) when all the attributes were used.

/pdf/predicting-diabetes-mellitus-with-machine-learning-1nnh9wy811.pdf

Predicting Diabetes Mellitus With Machine Learning Techniques.

MOTIVATION: A clear identification of the primary site of tumor is of great importance to the next targeted site-specific treatments and could efficiently improve patient‘s overall survival. Even though many classifiers based on gene expression had been proposed to predict the tumor primary, only a few studies focus on using DNA methylation (DNAm) profiles to develop classifiers, and none of them compares the performance of classifiers based on different profiles.
RESULTS: We introduced novel selection strategies to identify highly tissue-specific CpG sites and then used the random forest approach to construct the classifiers to predict the origin of tumors. We also compared the prediction performance by applying similar strategy on miRNA expression profiles. Our analysis indicated that these classifiers had an accuracy of 96.05% (Maximum–Relevance–Maximum–Distance: 90.02–99.99%) or 95.31% (principal component analysis: 79.82–99.91%) on independent DNAm datasets, and an overall accuracy of 91.30% (range 79.33–98.74%) on independent miRNA test sets for predicting tumor origin. This suggests that our feature selection methods are very effective to identify tissue-specific biomarkers and the classifiers we developed can efficiently predict the origin of tumors. We also developed a user-friendly webserver that helps users to predict the tumor origin by uploading miRNA expression or DNAm profile of their interests.
AVAILABILITY AND IMPLEMENTATION: The webserver, and relative data, code are accessible at http://server.malab.cn/MMCOP/.

/pdf/tumor-origin-detection-with-tissue-specific-mirna-and-dna-5dc2zjxb2y.pdf

Tumor origin detection with tissue-specific miRNA and DNA methylation markers.

Classical scoring functions have reached a plateau in their performance in virtual screening and binding affinity prediction. Recently, machine-learning scoring functions trained on protein-ligand complexes have shown great promise in small tailored studies. They have also raised controversy, specifically concerning model overfitting and applicability to novel targets. Here we provide a new ready-to-use scoring function (RF-Score-VS) trained on 15 426 active and 893 897 inactive molecules docked to a set of 102 targets. We use the full DUD-E data sets along with three docking tools, five classical and three machine-learning scoring functions for model building and performance assessment. Our results show RF-Score-VS can substantially improve virtual screening performance: RF-Score-VS top 1% provides 55.6% hit rate, whereas that of Vina only 16.2% (for smaller percent the difference is even more encouraging: RF-Score-VS top 0.1% achieves 88.6% hit rate for 27.5% using Vina). In addition, RF-Score-VS provides much better prediction of measured binding affinity than Vina (Pearson correlation of 0.56 and -0.18, respectively). Lastly, we test RF-Score-VS on an independent test set from the DEKOIS benchmark and observed comparable results. We provide full data sets to facilitate further research in this area (http://github.com/oddt/rfscorevs) as well as ready-to-use RF-Score-VS (http://github.com/oddt/rfscorevs_binary).

/pdf/performance-of-machine-learning-scoring-functions-in-3f3gnb3cab.pdf

Performance of machine-learning scoring functions in structure-based virtual screening.

The task of predicting the interactions between drugs and targets plays a key role in the process of drug discovery. There is a need to develop novel and efficient prediction approaches in order to avoid costly and laborious yet not-always-deterministic experiments to determine drug-target interactions (DTIs) by experiments alone. These approaches should be capable of identifying the potential DTIs in a timely manner. In this article, we describe the data required for the task of DTI prediction followed by a comprehensive catalog consisting of machine learning methods and databases, which have been proposed and utilized to predict DTIs. The advantages and disadvantages of each set of methods are also briefly discussed. Lastly, the challenges one may face in prediction of DTI using machine learning approaches are highlighted and we conclude by shedding some lights on important future research directions.

Machine learning approaches and databases for prediction of drug-target interaction: a survey paper.

A novel features ranking metric with application to scalable visual and bioinformatics data classification

An Empirical Study of Features Fusion Techniques for Protein-Protein Interaction Prediction

Selective ensemble learning is a method that selects a subset of diverse and accurate base models to generate stronger generalization ability. In this paper, we propose a selective ensemble learning algorithm called PTHS and a novel feature selection method called MSRD to solve the problem of high dimensionality. The algorithm PTHS uses a parallel optimization and hierarchical selection framework. The experimental result showed that MSRD is a suitable feature selection method for solving the problem of high dimensionality and that PTHS achieved better performance than other methods.

Jiancang Zeng

Papers

A novel features ranking metric with application to scalable visual and bioinformatics data classification

An Empirical Study of Features Fusion Techniques for Protein-Protein Interaction Prediction

Selective ensemble learning with parallel optimization and hierarchical selection