A novel CpG island methylation panel predicts survival in lung adenocarcinomas

Genome-wide promoter methylation profiling in a cellular model of melanoma progression reveals markers of malignancy and metastasis that predict melanoma survival

The epigenetic changes associated with melanoma progression to advanced and metastatic stages are still poorly understood. To shed light on the CpG methylation dynamics during melanoma development, we analyzed the methylome profiles of a four-stage cell line model of melanoma progression: non-tumorigenic melanocytes (melan-a), premalignant melanocytes (4C), non-metastatic melanoma cells (4C11−), and metastatic melanoma cells (4C11+). We identified 540 hypo- and 37 hypermethylated gene promoters that together characterized a malignancy signature, and 646 hypo- and 520 hypermethylated promoters that distinguished a metastasis signature. Differentially methylated genes from these signatures were correlated with overall survival using TCGA-SKCM methylation data. Moreover, multivariate Cox analyses with LASSO regularization identified panels of 33 and 31 CpGs, respectively, from the malignancy and metastasis signatures that predicted poor survival. We found a concordant relationship between DNA methylation and transcriptional levels for genes from the malignancy (Pyroxd2 and Ptgfrn) and metastasis (Arnt2, Igfbp4 and Ptprf) signatures, which were both also correlated with melanoma prognosis. Altogether, this study reveals novel CpGs methylation markers associated with malignancy and metastasis that collectively could improve the survival prediction of melanoma patients.

An Unbiased Predictive Model to Detect DNA Methylation Propensity of CpG Islands in the Human Genome

Background:

Epigenetic repression mechanisms play an important role in gene regulation, specifically in cancer development. In many cases, a CpG island’s (CGI) susceptibility or resistance to methylation is shown to be contributed by local DNA sequence features.

Objective:

To develop unbiased machine learning models–individually and combined for different biological features–that predict the methylation propensity of a CGI.

Methods:

We developed our model consisting of CGI sequence features on a dataset of 75 sequences (28 prone, 47 resistant) representing a genome-wide methylation structure. We tested our model on two independent datasets that are chromosome (132 sequences) and disease (70 sequences) specific.

Results:

We provided improvements in prediction accuracy over previous models. Our results indicate that combined features better predict the methylation propensity of a CGI (area under the curve (AUC) ~0.81). Our global methylation classifier performs well on independent datasets reaching an AUC of ~0.82 for the complete model and an AUC of ~0.88 for the model using select sequences that better represent their classes in the training set. We report certain de novo motifs and transcription factor binding site (TFBS) motifs that are consistently better in separating prone and resistant CGIs.

Conclusion:

Predictive models for the methylation propensity of CGIs lead to a better understanding of disease mechanisms and can be used to classify genes based on their tendency to contain methylation prone CGIs, which may lead to preventative treatment strategies. MATLAB® and Python™ scripts used for model building, prediction, and downstream analyses are available at https://github.com/dicleyalcin/methylProp_predictor.