23º SINAPE - Simpósio Nacional de Probabilidade e Estatística

Dados do Trabalho


Título

MLML2R: EFFICIENT SIMULTANEOUS MAXIMUM LIKELIHOOD ESTIMATION OF DNA METHYLATION AND HYDROXYMETHYLATION PROPORTIONS

Resumo

Single-base profiling of 5-methylcytosine (5-mC) and 5-hydroxymethylcytosine (5-hmC), requires combining data from DNA processing methods such as traditional (BS), oxidative (oxBS) or Tet-Assisted (TAB) bisulfite conversion. While other R packages provide 5-mC and 5-hmC maximum likelihood estimates (MLE) only for oxBS+BS combination, the R package MLML2R provides MLE also for TAB combinations. For combinations of any two of the methods, we derived the exact constrained MLE in analytical form, which greatly decreases computational time. For the three methods combination, we implemented the iterative method by Qu et al. (2013), and a non iterative approximation using Lagrange multipliers. The MLML2R package is freely available at https://CRAN. R- project.org/package=MLML2R.

Palavras-chave

DNA methylation, DNA hydroxymethylation, EM-algorithm, Maximum Likelihood Estimation, R package

Área

Estatística Aplicada em Agronomia e Biologia

Autores

Samara Flamini Kiihl, Maria José Martinez-Garrido, Arce Domingo-Relloso, Jose Bermudez, Maria Tellez-Plaza