EpiXcan Based Analysis

Introduction of EpiXcan Based Analysis

EpiXcan, which uses a Bayesian hierarchical model, tries to optimize the weaknesses of PrediXcan by applying epigenetic annotations to optimize the weights assigned to cis-eQTL and improve the predictability of gene expression levels. It is a novel method that increses prediction accuracy in transcriptome imputation by integrating epigenetic data to model the prior probability that a SNP affects transcription.

The principle of EpiXcan analysis is to integrate biological related data into a framework (such as leverages epigenetic annotation to inform transcriptomic imputation) to improve the performance of gene expression prediction. The specific analysis steps are as follows:

  • Firstly, based on a Bayesian hierarchical model that integrates epigenomic annotation and eQTL summary statistics for cis-SNPs (SNPs located ± 1 Mb from the transcription start site of the gene) to estimate SNP priors that reflect the likelihood of a SNP having a regulatory role in gene expression. Infer the effect of allele-specific variation.
  • Secondly, by employing a novel adaptive mapping approach to rescale the SNP priors to penalty factors.
  • Thirdly, use the genotype and penalty factors in the weighted elastic net to predict gene expression.
Fig 1. Comparison of gene-trait associations between EpiXcan and PrediXcan.

Fig 1. Comparison of gene-trait associations between EpiXcan and PrediXcan. (Zhang W, et al. 2019)

Advantages of EpiXcan Based Analysis

  • EpiXcan improves the average R2CVM, and improves the average R2PP, and the ratios of genes predicted more effectively.
  • EpiXcan uncovers more clinically relevant genes and molecular pathways.
  • Overall, EpiXcan has improved prediction performance compared to PrediXcan and identified more genes that can be used in TWAS.

Application Filed of EpiXcan

Transcriptome-wide association studies (TWAS) integrate gene expression data with common risk variation to identify gene-trait associations. Since TWAS is limited to genes that can be accurately predicted from genotype data, many downstream analyses are restricted. The analysis method of EpiXcan can improve the accuracy of prediction, thereby increasing the scope and ability of analysis. Overall, EpiXcan method utilizes epigenomic information to further improve prediction of transcriptomes and it provides a framework for TWASs, improving interrogation of traitassociated biological pathway involvement, and a platform for drug utilization and treatment development.

CD ComputaBio uses EpiXcan, which uses a Bayesian hierarchical model, to improve the accuracy of genes that predicted from tanscriptome-wide association studies data. With the help of Bayesian genetic models, EpiXcan has more power to detect significant genes, including novel and unique associations, which are indispensable for life and clinically significant. For tanscriptome-wide association studies data analysis, we provide different analysis method such as EpiXcan and PrediXcan according to your needs. In addition, with the help of artificial intelligence methods, we provide a one-stop biological data analysis service. You only need to provide the original data, and we will provide you with complete results reports and charts. Regarding data analysis, if you have any questions, please feel free to contact us. Sincerely at your service.


  • Zhang W, et al. Integrative transcriptome imputation reveals tissue-specific and shared biological mechanisms mediating susceptibility to complex traits[J]. Nature Communications. 2019.


Online Inquiry

CD ComputaBio

Copyright © 2024 CD ComputaBio Inc. All Rights Reserved.