Predicting ATP-Binding Cassette Transporters Using the Random Forest Method

Liu

2021

Front. Cell Dev. Biol.

ATP-binding cassette (ABC) transporters can promote cells to absorb nutrients and excrete harmful substances. It plays a vital role in the transmembrane transport of macromolecules. Therefore, the identification of ABC transporters is of great significance for the biological research. This paper will introduce a novel method called DeepRTCP. DeepRTCP uses the deep convolutional neural network and a feature combined of reduced amino acid alphabet based tripeptide composition and PSSM to recognize ABC transporters. We constructed a dataset named ABC_2020. It contains the latest ABC transporters downloaded from Uniprot. We performed 10-fold cross-validation on DeepRTCP, and the average accuracy of DeepRTCP was 95.96%. Compared with the start-of-the-art method for predicting ABC transporters, DeepRTCP improved the accuracy by 9.29%. It is anticipated that DeepRTCP can be used as an effective ABC transporter classifier which provides a reliable guidance for the research of ABC transporters.

Section: Resultsmentioning

confidence: 99%

DeepRTCP: Predicting ATP-Binding Cassette Transporters Based on 1-Dimensional Convolutional Network

Zhang

Liu

2021

Front. Cell Dev. Biol.

“…Han et al applied support vector machine and random forest methods to predict ion channels and their types from protein sequences [17]. Hou et al proposed a model combining 188D features with random forest to identify ABC transporters [15].…”

Section: Introductionmentioning

confidence: 99%

“…With the advent of the era of big data, machine learning (ML) techniques have been increasingly used as a powerful approach to identify important proteins in biology [ 14 ]. Although this method cannot replace biological experiments, it improves the accuracy of prediction and provides more clues for biological experiments [ 15 ]. There are many examples of protein identification using ML approaches, and most of them show good predictive performance.…”

Section: Introductionmentioning

confidence: 99%

See 1 more Smart Citation

Prediction of prokaryotic transposases from protein features with machine learning approaches

et al. 2021

Microbial Genomics

Identification of prokaryotic transposases (Tnps) not only gives insight into the spread of antibiotic resistance and virulence but the process of DNA movement. This study aimed to develop a classifier for predicting Tnps in bacteria and archaea using machine learning (ML) approaches. We extracted a total of 2751 protein features from the training dataset including 14852 Tnps and 14852 controls, and selected 75 features as predictive signatures using the combined mutual information and least absolute shrinkage and selection operator algorithms. By aggregating these signatures, an ensemble classifier that integrated a collection of individual ML-based classifiers, was developed to identify Tnps. Further validation revealed that this classifier achieved good performance with an average AUC of 0.955, and met or exceeded other common methods. Based on this ensemble classifier, a stand-alone command-line tool designated TnpDiscovery was established to maximize the convenience for bioinformaticians and experimental researchers toward Tnp prediction. This study demonstrates the effectiveness of ML approaches in identifying Tnps, facilitating the discovery of novel Tnps in the future.

“…Endometrial carcinoma (EC) is a common malignancy of the female reproductive system, the incidence of which is increasing [1]. EC is a heterogeneous tumor, and the prognosis of patients is closely related to the tumor grade and stage.…”

Section: Introductionmentioning

confidence: 99%

Significance of PTEN Mutational Status Associated Gene Signature in the Progression and Prognosis of Endometrial Carcinoma

Liu

et al. 2021

Preprint

BackgroundPTEN mutation had been reported to be involved in the development and prognosis of endometrial carcinoma (EC). However, a prognostic genes signature associated with PTEN mutational status has not been developed. In this study, we aim to conduct a PTEN mutation associated prognostic genes signature for EC.MethodsWe obtained the simple nucleotide variation and transcriptome profiling data from The Cancer Genome Atlas database as training data. Lasso Cox regression algorithm was used to establish PTEN mutation associated prognostic genes signature. The overall survival rate of the high-risk and low-risk groups was determined by Kaplan-Meier (K-M) method. The accuracy of risk score prediction was tested by ROC curve.ResultsK-M curves revealed that the EC patients with PTEN mutation have favorable survival outcome. Differential expression analysis between the EC patients with PTEN mutation and PTEN wild identified 224 differential expression genes (DEGs). Eighty-four DEGs with prognostic value was fitted into least absolute shrinkage and selection operator (LASSO)–Cox analysis and a seven PTEN mutation associated prognostic genes signature with robust prognostic ability was constructed, which was successfully validated in the other two datasets from cBioPortal database as well as 60 clinical specimens. Furthermore, the PTEN mutation associated prognostic genes signature had been proved to be an independent prognostic predictor for EC. Remarkably, the EC patients in high-risk group were characterized with higher stages and grades as well as lower tumor mutation burden of EC, with poor survival outcome. Collectively, the PTEN mutation associated prognostic genes signature was a favorable prognostic biomarker for EC.ConclusionIn summary, we conducted and validated a prognostic predictor for EC associated with PTEN mutational status, which may be used as favorable prognostic biomarkers and therapeutic targets for EC.