Dataset Information

Prediction and prioritization of autism-associated long non-coding RNAs using gene expression and sequence features.

ABSTRACT:

Background

Autism spectrum disorders (ASD) refer to a range of neurodevelopmental conditions, which are genetically complex and heterogeneous with most of the genetic risk factors also found in the unaffected general population. Although all the currently known ASD risk genes code for proteins, long non-coding RNAs (lncRNAs) as essential regulators of gene expression have been implicated in ASD. Some lncRNAs show altered expression levels in autistic brains, but their roles in ASD pathogenesis are still unclear.

Results

In this study, we have developed a new machine learning approach to predict candidate lncRNAs associated with ASD. Particularly, the knowledge learnt from protein-coding ASD risk genes was transferred to the prediction and prioritization of ASD-associated lncRNAs. Both developmental brain gene expression data and transcript sequence were found to contain relevant information for ASD risk gene prediction. During the pre-training phase of model construction, an autoencoder network was implemented for a representation learning of the gene expression data, and a random-forest-based feature selection was applied to the transcript-sequence-derived k-mers. Our models, including logistic regression, support vector machine and random forest, showed robust performance based on tenfold cross-validations as well as candidate prioritization with hypothetical loci. We then utilized the models to predict and prioritize a list of candidate lncRNAs, including some reported to be cis-regulators of known ASD risk genes, for further investigation.

Conclusions

Our results suggest that ASD risk genes can be accurately predicted using developmental brain gene expression data and transcript sequence features, and the models may provide useful information for functional characterization of the candidate lncRNAs associated with ASD.

SUBMITTER: Wang J

PROVIDER: S-EPMC7648398 | biostudies-literature | 2020 Nov

REPOSITORIES: biostudies-literature

ACCESS DATA

Publications

Prediction and prioritization of autism-associated long non-coding RNAs using gene expression and sequence features.

Wang Jun J Wang Liangjiang L

BMC bioinformatics 20201107 1

<h4>Background</h4>Autism spectrum disorders (ASD) refer to a range of neurodevelopmental conditions, which are genetically complex and heterogeneous with most of the genetic risk factors also found in the unaffected general population. Although all the currently known ASD risk genes code for proteins, long non-coding RNAs (lncRNAs) as essential regulators of gene expression have been implicated in ASD. Some lncRNAs show altered expression levels in autistic brains, but their roles in ASD pathog ...[more]

PMID: 33160303

Similar Datasets

Project description:Genetic and environmental factors, alone or in combination, contribute to the pathogenesis of autism spectrum disorder (ASD). Although many protein-coding genes have now been identified as disease risk genes for ASD, a detailed illustration of long non-coding RNAs (lncRNAs) associated with ASD remains elusive. In this study, we first identified ASD-related lncRNAs based on genomic variant data of individuals with ASD from a twin study. In total, 532 ASD-related lncRNAs were identified, and 86.7% of these ASD-related lncRNAs were further validated by an independent copy number variant (CNV) dataset. Then, the functions and associated biological pathways of ASD-related lncRNAs were explored by enrichment analysis of their three different types of functional neighbor genes (i.e., genomic neighbors, competing endogenous RNA (ceRNA) neighbors, and gene co-expression neighbors in the cortex). The results have shown that most of the functional neighbor genes of ASD-related lncRNAs were enriched in nervous system development, inflammatory responses, and transcriptional regulation. Moreover, we explored the differential functions of ASD-related lncRNAs in distinct brain regions by using gene co-expression network analysis based on tissue-specific gene expression profiles. As a set, ASD-related lncRNAs were mainly associated with nervous system development and dopaminergic synapse in the cortex, but associated with transcriptional regulation in the cerebellum. In addition, a functional network analysis was conducted for the highly reliable functional neighbor genes of ASD-related lncRNAs. We found that all the highly reliable functional neighbor genes were connected in a single functional network, which provided additional clues for the action mechanisms of ASD-related lncRNAs. Finally, we predicted several potential drugs based on the enrichment of drug-induced pathway sets in the ASD-altered biological pathway list. Among these drugs, several (e.g., amoxapine, piperine, and diflunisal) were partly supported by the previous reports. In conclusion, ASD-related lncRNAs participated in the pathogenesis of ASD through various known biological pathways, which may be differential in distinct brain regions. Detailed investigation into ASD-related lncRNAs can provide clues for developing potential ASD diagnosis biomarkers and therapy.

Project description:BackgroundWilms tumor (WT) is the most frequent malignancy of the kidney in children, and a subset of patients remains with a poor prognosis. This study aimed to identify key long non-coding RNAs (lncRNAs) related to prognosis and establish a genomic-clinicopathologic nomogram to predict survival in children with WT.MethodsClinical data of 124 WT patients and the relevant RNA sequencing data including lncRNAs expression signature of primary WT samples were obtained from the Therapeutically Applicable Research to Generate Effective Treatment (TARGET) Data Matrix. Then, lncRNAs associated with overall survival (OS) were identified through univariate Cox, least absolute shrinkage and selection operator (LASSO), and multivariate Cox regression analyses. The risk scores of 124 participants were calculated, and survival analyses were performed between low- and high-risk groups. A genomic-clinicopathologic nomogram was then developed and evaluated by time-dependent receiver operating characteristic (ROC) curves, including the area under the curve (AUC), calibration curve, and decision curve analysis. Subsequently, bioinformatics analyses were performed to explore the potential molecular mechanisms that affect the prognosis of WT. The package "DESeq2" was used to identify differentially expressed protein-coding genes (DEPCGs) between groups. Gene Set Enrichment Analysis (GSEA) was applied to explore the differences in pathways enrichment. The analytical tools CIBERSORTx and ESTIMATE were used to investigate the discrepancies of the immune microenvironment.ResultsA total of 10 lncRNAs were selected as independent predictors associated with OS (P<0.05). Participants in the high-risk group had a significantly worse OS and event-free survival (EFS) than those in the low-risk group (P<2E-16 and P=2.03E-04, respectively). The risk score and 3 clinicopathological features (gender, cooperative group protocol, and stage) were identified to construct the nomogram (combined model) (P=5.11E-17). The combined model (1-year AUC: 0.9272, 3-year AUC: 0.9428, 5-year AUC: 0.9259) and risk score model (1-year AUC: 0.9285, 3-year AUC: 0.9399, 5-year AUC: 0.9266) displayed higher predictive accuracy than that of the other models. Subsequently, 105 DEPCGs were identified. The GSEA revealed 4 significant pathways. Analysis with CIBERSORTx demonstrated that monocytes, macrophages M1, activated dendritic cells, and resting mast cells had significant infiltration differences between groups.ConclusionsThis study constructed a genomic-clinicopathologic nomogram, which might present a novel and efficient method for treating patients with WT.

Dataset Information

Prediction and prioritization of autism-associated long non-coding RNAs using gene expression and sequence features.

Background

Results

Conclusions

Publications

Prediction and prioritization of autism-associated long non-coding RNAs using gene expression and sequence features.

Similar Datasets

OmicsDI is part of the ELIXIR infrastructure

Tweets