Skip to main content

Table 4 The bisulfite conversion model

From: epiG: statistical inference and profiling of DNA methylation from whole-genome bisulfite sequencing data

 

Epigenotype g

 

Z

C

G

A

T

\(\overset {\text {me}}{\mathsf {C}}\)

\(\underset {\text {me}}{\mathsf {G}}\)

C

\(\hat \rho \alpha \)

0

0

0

\(1 - \hat \rho \beta \)

0

G

0

ρ α

0

0

0

1−ρ β

A

0

1−ρ α

1

0

0

ρ β

T

\(1 - \hat \rho \alpha \)

0

0

1

\(\hat \rho \beta \)

0

  1. The probability P(Z=zg,r) of Z given the epigenotype g and the strand r. Here α is the rate of failed bisulfite conversion and β is the rate of inappropriately converted nucleotides. The parameter ρ is 0 if r=fwd, 1 if r=rev, and \(\hat \rho =1-\rho \). For example, P(Z=Cg=C,r=fwd)=α