ABSTRACT
In this review, we provide a comprehensive overview of the different computational tools that have been published for the deconvolution of bulk DNA methylation (DNAm) data. Here, deconvolution refers to the estimation of cell-type proportions that constitute a mixed sample. The paper reviews and compares 25 deconvolution methods (supervised, unsupervised or hybrid) developed between 2012 and 2023 and compares the strengths and limitations of each approach. Moreover, in this study, we describe the impact of the platform used for the generation of methylation data (including microarrays and sequencing), the applied data pre-processing steps and the used reference dataset on the deconvolution performance. Next to reference-based methods, we also examine methods that require only partial reference datasets or require no reference set at all. In this review, we provide guidelines for the use of specific methods dependent on the DNA methylation data type and data availability.
Subject(s)
Computational Biology , DNA Methylation , Humans , Computational Biology/methods , DNA/genetics , AlgorithmsABSTRACT
Introduction: Diffuse large B-cell lymphoma (DLBCL) and primary mediastinal B-cell lymphoma (PMBCL) are aggressive histological subtypes of non-Hodgkin's lymphoma. Improved understanding of the underlying molecular pathogenesis has led to new classification and risk stratification tools, including the development of cell-free biomarkers through liquid biopsies. The goal of this study was to investigate cell-free RNA (cfRNA) biomarkers in DLBCL and PMBCL patients. Materials and methods: Blood plasma samples (n=168) and matched diagnostic formalin-fixed paraffin-embedded (FFPE) tissue samples (n=69) of DLBCL patients, PMBCL patients and healthy controls were collected between 2016-2021. Plasma samples were collected at diagnosis, at interim evaluation, after treatment, and in case of refractory or relapsed disease. RNA was extracted from 200 µl plasma using the miRNeasy serum/plasma kit and from FFPE tissue using the miRNeasy FFPE kit. RNA was subsequently sequenced on a NovaSeq 6000 instrument using the SMARTer Stranded Total RNA-seq pico v3 library preparation kit. Results: Higher cfRNA concentrations were demonstrated in lymphoma patients compared to healthy controls. A large number of differentially abundant genes were identified between the cell-free transcriptomes of DLBCL patients, PMBCL patients, and healthy controls. Overlap analyses with matched FFPE samples showed that blood plasma has a unique transcriptomic profile that significantly differs from that of the tumor tissue. As a good concordance between tissue-derived gene expression and the immunohistochemistry Hans algorithm for cell-of-origin (COO) classification was demonstrated in the FFPE samples, but not in the plasma samples, a 64-gene cfRNA classifier was developed that can accurately determine COO in plasma. High plasma levels of a 9-gene signature (BECN1, PRKCB, COPA, TSC22D3, MAP2K3, UQCRHL, PTMAP4, EHD1, NAP1L1 pseudogene) and a 5-gene signature (FTH1P7, PTMAP4, ATF4, FTH1P8, ARMC7) were significantly associated with inferior progression-free and overall survival in DLBCL patients, respectively, independent of the NCCN-IPI score. Conclusion: Total RNA sequencing of blood plasma samples allows the analysis of the cell-free transcriptome in DLBCL and PMBCL patients and demonstrates its unexplored potential in identifying diagnostic, cell-of-origin, and prognostic cfRNA biomarkers.
ABSTRACT
Episignatures are popular tools for the diagnosis of rare neurodevelopmental disorders. They are commonly based on a set of differentially methylated CpGs used in combination with a support vector machine model. DNA methylation (DNAm) data often include missing values due to changes in data generation technology and batch effects. While many normalization methods exist for DNAm data, their impact on episignature performance have never been assessed. In addition, technologies to quantify DNAm evolve quickly and this may lead to poor transposition of existing episignatures generated on deprecated array versions to new ones. Indeed, probe removal between array versions, technologies or during preprocessing leads to missing values. Thus, the effect of missing data on episignature performance must also be carefully evaluated and addressed through imputation or an innovative approach to episignatures design. In this paper, we used data from patients suffering from Kabuki and Sotos syndrome to evaluate the influence of normalization methods, classification models and missing data on the prediction performances of two existing episignatures. We compare how six popular normalization methods for methylarray data affect episignature classification performances in Kabuki and Sotos syndromes and provide best practice suggestions when building new episignatures. In this setting, we show that Illumina, Noob or Funnorm normalization methods achieved higher classification performances on the testing sets compared to Quantile, Raw and Swan normalization methods. We further show that penalized logistic regression and support vector machines perform best in the classification of Kabuki and Sotos syndrome patients. Then, we describe a new paradigm to build episignatures based on the detection of differentially methylated regions (DMRs) and evaluate their performance compared to classical differentially methylated cytosines (DMCs)-based episignatures in the presence of missing data. We show that the performance of classical DMC-based episignatures suffers from the presence of missing data more than the DMR-based approach. We present a comprehensive evaluation of how the normalization of DNA methylation data affects episignature performance, using three popular classification models. We further evaluate how missing data affect those models' predictions. Finally, we propose a novel methodology to develop episignatures based on differentially methylated regions identification and show how this method slightly outperforms classical episignatures in the presence of missing data.
