InfiniumPurify-R
InfiniumPurify-R adjusts for tumor purity in DNA methylation analyses to estimate tumor purity from Infinium 450k array data and detect differentially methylated CpG sites while accounting for purity-associated confounding.
Key Features:
- Tumor Purity Estimation: The function getPurity estimates tumor purity from DNA methylation data generated on the Infinium 450k array.
- Differential Methylation Analysis: The InfiniumDMC function employs a generalized least squares model to adjust for tumor purity when identifying differentially methylated CpG sites between groups with distinct genotypes or phenotypes and uneven purity distributions.
- Sample Clustering: The InfiniumClust function performs clustering of tumor samples while accounting for estimated or user-provided tumor purities to support subtype clustering and subclonal deconvolution.
- Experimental Design Flexibility: Applicable to analyses with or without normal controls and to scenarios with different sources of normal tissue contamination.
- Performance Evaluation: Validation was performed on simulated datasets and TCGA breast cancer samples and compared against methods such as minfi, limma, and limma corrected by tumor purity across varying differential methylation thresholds, sample sizes, and mean purity deviations.
Scientific Applications:
- Epigenetic mechanism discovery: Detection of differentially methylated CpG sites to investigate epigenetic changes underlying tumorigenesis.
- Biomarker identification and subtyping: Identification of methylation biomarkers for cancer subtyping and stratification.
- Mitigation of sample heterogeneity: Accounting for tumor purity to reduce confounding from normal tissue contamination in methylation studies, supporting analyses for personalized medicine.
Methodology:
Computational steps explicitly include tumor purity estimation from Infinium 450k methylation data via getPurity, differential methylation testing using a generalized least squares model in InfiniumDMC, and clustering that accounts for estimated or user-provided tumor purities in InfiniumClust, with applicability to designs with or without normal controls and different sources of normal tissue contamination.
Topics
Details
- Tool Type:
- library
- Programming Languages:
- R
- Added:
- 1/14/2020
- Last Updated:
- 1/14/2021
Operations
Publications
Zhang W, Li Z, Wei N, Wu H, Zheng X. Detection of differentially methylated CpG sites between tumor samples with uneven tumor purities. Bioinformatics. 2019;36(7):2017-2024. doi:10.1093/bioinformatics/btz885. PMID:31769783.