Model | Additive Gaussian noise is added to the data. |
μ
| Mean of the additive Gaussian noise. Noise is drawn from N (μ, α2) |
α
2
| Variance of the additive Gaussian noise. |
SNR EM: | |
Model | Additive Gaussian noise is added to the data with given signal-to-noise ratio. |
μ
| Mean of the additive Gaussian noise. |
SNR | Signal-to-noise ratio after the noise is added. |
Dror EM [7]: | |
Model | y = g * (x
i
* x) + f + ε |
, | Binding efficiency of each probe x
i
is drawn from Gaussian distribution N (, ). |
μ
f
, | Gene specific bias f is drawn from Gaussian distribution N (μ
f
, ). |
α
ε
, β
ε
| Gene and chip specific error ε is drawn from Laplace distribution L (α
ε
, β
ε
). |
μ
g
, | Multiplicative gene and chip specific noise g is drawn from log-normal distribution LN (μ
g
, ). |
Hartemink EM [9]: | |
Model | In log scale y = x + ρ
j
+ ε
ij
|
, | Chip specific bias ρ
j
is drawn from Gaussian distribution N (, ). |
| Gene and chip specific error ε
ij
is drawn from Gaussian distribution N (0, ). |
Hierarchical EM [6]: | |
Model | In log scale y = X + ε, X = x + g
i
+ c
j
, + r
ij
+ b
ijk
|
| Independent random noise ε is drawn from zero mean Gaussian distribution N (0, ). |
| Gene specific noise g
i
is drawn from zero mean Gaussian distribution.N (0, ). |
| Chip specific noise C
j
is drawn from zero mean Gaussian distribution N (0, ). |
| Gene and chip specific noise r
ij
is drawn from zero mean Gaussian distribution N (0, ). |
| Gene, chip and biological sample specific noise b
ijk
is drawn from zero mean Gaussian distribution N (0, ). |
Rocke EM [8]: | |
Model | y = α + xen + ε |
| Multiplicative noise n is drawn from zero mean Gaussian distribution N (0, ). |
| Additive independent noise ε is drawn from zero mean Gaussian distribution N (0, ). |
μ
α
, | Background noise (bias) α is drawn from Gaussian distribution N (μ
α
, ). |
Hein EM [11]: | |
Model | PM
ijkp
~ N(S
ijkp
+ H
ijkp
, ), MM
ijkp
~ N (φS
ijkp
+ H
ijkp
, ), where PM refers to perfect match and MM to mismatch probe. |
a
k
, | True expression signal log(S
ijkp
+ 1) is drawn from truncated (realization always ≥ 0) Gaussian distribution TN (x, ), where variance is drawn from Gaussian distribution N (a
k
, ) and x is the underlying expression value. |
μ
λ
, , α
η
, β
η
| Hybridization error term log(H
ijkp
+ 1) is drawn from truncated Gaussian distribution TN (λ
jk
, ). Parameter λ
jk
is drawn from Gaussian distribution N (μ
λ
, ) and is drawn from gamma distribution Γ-1(α
τ
, β
τ
). |
α
τ
, β
τ
| Variance is drawn from gamma distribution Γ-1(α
τ
, β
τ
). |
φ
| Fractional binding φ can be selected from interval [0, 1]. |