A similarity-based method for prediction of drug side effects with heterogeneous information

Xian, Zhao; Chen, Lei; Lü, Jing

doi:10.1016/j.mbs.2018.09.010

Cited by 142 publications

(151 citation statements)

References 51 publications

Supporting

Mentioning

150

Contrasting

Unclassified

Order By: Relevance

“…In the first stage, all features were ranked by a powerful feature selection method, i.e., the mRMR method [ 25 ]. This method has been widely used to rank features in several studies [ 71 , 72 , 73 , 74 , 75 , 76 , 77 , 78 , 79 , 80 ]. In this method, two criteria, namely, (1) Max-Relevance and (2) Min-Redundancy, were adopted to rank features singly or simultaneously.…”

Section: Methodsmentioning

confidence: 99%

Tissue Expression Difference between mRNAs and lncRNAs

Lei

Zhang

Pan

et al. 2018

IJMS

Self Cite

View full text Add to dashboard Cite

Messenger RNA (mRNA) and long noncoding RNA (lncRNA) are two main subgroups of RNAs participating in transcription regulation. With the development of next generation sequencing, increasing lncRNAs are identified. Many hidden functions of lncRNAs are also revealed. However, the differences in lncRNAs and mRNAs are still unclear. For example, we need to determine whether lncRNAs have stronger tissue specificity than mRNAs and which tissues have more lncRNAs expressed. To investigate such tissue expression difference between mRNAs and lncRNAs, we encoded 9339 lncRNAs and 14,294 mRNAs with 71 expression features, including 69 maximum expression features for 69 types of cells, one feature for the maximum expression in all cells, and one expression specificity feature that was measured as Chao-Shen-corrected Shannon’s entropy. With advanced feature selection methods, such as maximum relevance minimum redundancy, incremental feature selection methods, and random forest algorithm, 13 features presented the dissimilarity of lncRNAs and mRNAs. The 11 cell subtype features indicated which cell types of the lncRNAs and mRNAs had the largest expression difference. Such cell subtypes may be the potential cell models for lncRNA identification and function investigation. The expression specificity feature suggested that the cell types to express mRNAs and lncRNAs were different. The maximum expression feature suggested that the maximum expression levels of mRNAs and lncRNAs were different. In addition, the rule learning algorithm, repeated incremental pruning to produce error reduction algorithm, was also employed to produce effective classification rules for classifying lncRNAs and mRNAs, which gave competitive results compared with random forest and could give a clearer picture of different expression patterns between lncRNAs and mRNAs. Results not only revealed the heterogeneous expression pattern of lncRNA and mRNA, but also gave rise to the development of a new tool to identify the potential biological functions of such RNA subgroups.

show abstract

Section: Methodsmentioning

confidence: 99%

Tissue Expression Difference between mRNAs and lncRNAs

Lei

Zhang

Pan

et al. 2018

IJMS

Self Cite

View full text Add to dashboard Cite

show abstract

“…For this type of dataset, the overall accuracy cannot correctly indicate the quality of predicted results because it is highly related to the accuracy of the largest class. For binary classification, Matthews correlation coefficient (MCC) [ 40 , 41 , 42 , 43 ] is regarded as a balanced measure, even if the classes are of very different sizes. In this study, we employed its multiclass version [ 44 ], which was proposed by Gorodkin, to evaluate the prediction performance using ten-fold cross-validation [ 31 , 45 , 46 , 47 ].…”

Section: Methodsmentioning

confidence: 99%

Identification of the Gene Expression Rules That Define the Subtypes in Glioma

Cai

Zhang

et al. 2018

JCM

Self Cite

View full text Add to dashboard Cite

As a common brain cancer derived from glial cells, gliomas have three subtypes: glioblastoma, diffuse astrocytoma, and anaplastic astrocytoma. The subtypes have distinctive clinical features but are closely related to each other. A glioblastoma can be derived from the early stage of diffuse astrocytoma, which can be transformed into anaplastic astrocytoma. Due to the complexity of these dynamic processes, single-cell gene expression profiles are extremely helpful to understand what defines these subtypes. We analyzed the single-cell gene expression profiles of 5057 cells of anaplastic astrocytoma tissues, 261 cells of diffuse astrocytoma tissues, and 1023 cells of glioblastoma tissues with advanced machine learning methods. In detail, a powerful feature selection method, Monte Carlo feature selection (MCFS) method, was adopted to analyze the gene expression profiles of cells, resulting in a feature list. Then, the incremental feature selection (IFS) method was applied to the obtained feature list, with the help of support vector machine (SVM), to extract key features (genes) and construct an optimal SVM classifier. Several key biomarker genes, such as IGFBP2, IGF2BP3, PRDX1, NOV, NEFL, HOXA10, GNG12, SPRY4, and BCL11A, were identified. In addition, the underlying rules of classifying the three subtypes were produced by Johnson reducer algorithm. We found that in diffuse astrocytoma, PRDX1 is highly expressed, and in glioblastoma, the expression level of PRDX1 is low. These rules revealed the difference among the three subtypes, and how they are formed and transformed. These genes are not only biomarkers for glioma subtypes, but also drug targets that may switch the clinical features or even reverse the tumor progression.

show abstract

“…In this study, we showed the results of RF in comparison with the results obtained in previous studies [37,38]. ey adopted RF as their final model.…”

Section: Comparison With Previous Studiesmentioning

confidence: 73%

“…We compared our results with those of two previous studies by Zhao et al because they too treated a drug-side effect pair as a sample, like we did in this study [37,38]. Two other previous studies used the fingerprint, chemical structure, ATC code, literature association, and target as features to predict side effects and applied diverse machine learning algorithms such as RF, nearest neighbor, dagging, and support vector machine.…”

Section: Comparison With Previous Studiesmentioning

confidence: 97%

Prediction of Side Effects Using Comprehensive Similarity Measures

Seo

Lee

Kim

et al. 2020

BioMed Research International

View full text Add to dashboard Cite

Identifying the potential side effects of drugs is crucial in clinical trials in the pharmaceutical industry. e existing side effect prediction methods mainly focus on the chemical and biological properties of drugs. is study proposes a method that uses diverse information such as drug-drug interactions from DrugBank, drug-drug interactions from network, single nucleotide polymorphisms, and side effect anatomical hierarchy as well as chemical structures, indications, and targets. e proposed method is based on the assumption that properties used in drug repositioning studies could be utilized to predict side effects because the phenotypic expression of a side effect is similar to that of the disease. e prediction results using the proposed method showed a 3.5% improvement in the area under the curve (AUC) over that obtained when only chemical, indication, and target features were used. e random forest model delivered outstanding results for all combinations of feature types. Finally, after identifying candidate side effects of drugs using the proposed method, the following four popular drugs were discussed: (1) dasatinib, (2) sitagliptin, (3) vorinostat, and (4) clonidine.

show abstract

A similarity-based method for prediction of drug side effects with heterogeneous information

Cited by 142 publications

References 51 publications

Tissue Expression Difference between mRNAs and lncRNAs

Tissue Expression Difference between mRNAs and lncRNAs

Identification of the Gene Expression Rules That Define the Subtypes in Glioma

Prediction of Side Effects Using Comprehensive Similarity Measures

Contact Info

Product

Resources

About