Comparison of pre-processing methods for Infinium HumanMethylation450 BeadChip array

Shiah, Yu-Jia; Fraser, Michael; Bristow, Robert G.; Boutros, Paul C.

doi:10.1093/bioinformatics/btx372

Cited by 10 publications

(5 citation statements)

References 33 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…Shiah et al compared 11 preprocessing methods on a large 450K prostate cancer dataset and considered technical replicate variances and differences, within-batch clustering and inter-array correlations. They found Dasen and noob to be preprocessing methods which minimised technical differences [ 44 ]. Liu and Siegmund [ 45 ] compared 9 methods across 4 datasets and found the combinations of noob with SWAN or BMIQ were optimal within-array methods and functional normalisation, subset quantile normalisation (SQN) and Dasen optimal between-array methods.…”

Section: Discussionmentioning

confidence: 99%

Batch-effect detection, correction and characterisation in Illumina HumanMethylation450 and MethylationEPIC BeadChip array data

et al. 2022

View full text Add to dashboard Cite

Background Genomic technologies can be subject to significant batch-effects which are known to reduce experimental power and to potentially create false positive results. The Illumina Infinium Methylation BeadChip is a popular technology choice for epigenome-wide association studies (EWAS), but presently, little is known about the nature of batch-effects on these designs. Given the subtlety of biological phenotypes in many EWAS, control for batch-effects should be a consideration. Results Using the batch-effect removal approaches in the ComBat and Harman software, we examined two in-house datasets and compared results with three large publicly available datasets, (1214 HumanMethylation450 and 1094 MethylationEPIC BeadChips in total), and find that despite various forms of preprocessing, some batch-effects persist. This residual batch-effect is associated with the day of processing, the individual glass slide and the position of the array on the slide. Consistently across all datasets, 4649 probes required high amounts of correction. To understand the impact of this set to EWAS studies, we explored the literature and found three instances where persistently batch-effect prone probes have been reported in abstracts as key sites of differential methylation. As well as batch-effect susceptible probes, we also discover a set of probes which are erroneously corrected. We provide batch-effect workflows for Infinium Methylation data and provide reference matrices of batch-effect prone and erroneously corrected features across the five datasets spanning regionally diverse populations and three commonly collected biosamples (blood, buccal and saliva). Conclusions Batch-effects are ever present, even in high-quality data, and a strategy to deal with them should be part of experimental design, particularly for EWAS. Batch-effect removal tools are useful to reduce technical variance in Infinium Methylation data, but they need to be applied with care and make use of post hoc diagnostic measures.

show abstract

Section: Discussionmentioning

confidence: 99%

Batch-effect detection, correction and characterisation in Illumina HumanMethylation450 and MethylationEPIC BeadChip array data

et al. 2022

View full text Add to dashboard Cite

show abstract

“…To further evaluate the generalizability of this ensemble method, we executed the workflow on a prostate cancer methylation preprocessing dataset [ 33 ]. This set consists of the raw methylation values along with data from 11 preprocessing methods.…”

Section: Resultsmentioning

confidence: 99%

“…All samples incorporated in the analysis were surgical specimens taken prior to any treatment. To verify the ensemble method can be effective in other data types, a prostate cancer methylation preprocessing dataset containing 310 samples normalized using 11 different strategies was used [ 33 ].…”

Section: Methodsmentioning

confidence: 99%

Prediction of early breast cancer patient survival using ensembles of hypoxia signatures

et al. 2018

View full text Add to dashboard Cite

BackgroundBiomarkers are a key component of precision medicine. However, full clinical integration of biomarkers has been met with challenges, partly attributed to analytical difficulties. It has been shown that biomarker reproducibility is susceptible to data preprocessing approaches. Here, we systematically evaluated machine-learning ensembles of preprocessing methods as a general strategy to improve biomarker performance for prediction of survival from early breast cancer.ResultsWe risk stratified breast cancer patients into either low-risk or high-risk groups based on four published hypoxia signatures (Buffa, Winter, Hu, and Sorensen), using 24 different preprocessing approaches for microarray normalization. The 24 binary risk profiles determined for each hypoxia signature were combined using a random forest to evaluate the efficacy of a preprocessing ensemble classifier. We demonstrate that the best way of merging preprocessing methods varies from signature to signature, and that there is likely no ‘best’ preprocessing pipeline that is universal across datasets, highlighting the need to evaluate ensembles of preprocessing algorithms. Further, we developed novel signatures for each preprocessing method and the risk classifications from each were incorporated in a meta-random forest model. Interestingly, the classification of these biomarkers and its ensemble show striking consistency, demonstrating that similar intrinsic biological information are being faithfully represented. As such, these classification patterns further confirm that there is a subset of patients whose prognosis is consistently challenging to predict.ConclusionsPerformance of different prognostic signatures varies with pre-processing method. A simple classifier by unanimous voting of classifications is a reliable way of improving on single preprocessing methods. Future signatures will likely require integration of intrinsic and extrinsic clinico-pathological variables to better predict disease-related outcomes.

show abstract

“…Seven normalization procedures were selected based on consistent demonstration of their high performance in the literature [ 48 , 78 – 80 ]: functional, functional + noob, beta-mixture quantile (BMIQ), BMIQ + noob, dasen, dasen + noob, and noob normalization alone. We computed the Spearman correlation coefficient (rho) and root-mean square errors (RMSE) of each sample’s raw versus normalized data.…”

Section: Methodsmentioning

confidence: 99%

Who’s afraid of the X? Incorporating the X and Y chromosomes into the analysis of DNA methylation array data

Inkster

Wong

Matthews

et al. 2023

Epigenetics & Chromatin

View full text Add to dashboard Cite

Background Many human disease phenotypes manifest differently by sex, making the development of methods for incorporating X and Y-chromosome data into analyses vital. Unfortunately, X and Y chromosome data are frequently excluded from large-scale analyses of the human genome and epigenome due to analytical complexity associated with sex chromosome dosage differences between XX and XY individuals, and the impact of X-chromosome inactivation (XCI) on the epigenome. As such, little attention has been given to considering the methods by which sex chromosome data may be included in analyses of DNA methylation (DNAme) array data. Results With Illumina Infinium HumanMethylation450 DNAme array data from 634 placental samples, we investigated the effects of probe filtering, normalization, and batch correction on DNAme data from the X and Y chromosomes. Processing steps were evaluated in both mixed-sex and sex-stratified subsets of the analysis cohort to identify whether including both sexes impacted processing results. We found that identification of probes that have a high detection p-value, or that are non-variable, should be performed in sex-stratified data subsets to avoid over- and under-estimation of the quantity of probes eligible for removal, respectively. All normalization techniques investigated returned X and Y DNAme data that were highly correlated with the raw data from the same samples. We found no difference in batch correction results after application to mixed-sex or sex-stratified cohorts. Additionally, we identify two analytical methods suitable for XY chromosome data, the choice between which should be guided by the research question of interest, and we performed a proof-of-concept analysis studying differential DNAme on the X and Y chromosome in the context of placental acute chorioamnionitis. Finally, we provide an annotation of probe types that may be desirable to filter in X and Y chromosome analyses, including probes in repetitive elements, the X-transposed region, and cancer-testis gene promoters. Conclusion While there may be no single “best” approach for analyzing DNAme array data from the X and Y chromosome, analysts must consider key factors during processing and analysis of sex chromosome data to accommodate the underlying biology of these chromosomes, and the technical limitations of DNA methylation arrays.

show abstract

Comparison of pre-processing methods for Infinium HumanMethylation450 BeadChip array

Abstract: Supplementary data are available at Bioinformatics online.

Cited by 10 publications

References 33 publications

Batch-effect detection, correction and characterisation in Illumina HumanMethylation450 and MethylationEPIC BeadChip array data

Batch-effect detection, correction and characterisation in Illumina HumanMethylation450 and MethylationEPIC BeadChip array data

Prediction of early breast cancer patient survival using ensembles of hypoxia signatures

Who’s afraid of the X? Incorporating the X and Y chromosomes into the analysis of DNA methylation array data

Contact Info

Product

Resources

About