Skip to main content

Table 2 Sampling efficiency in the population parameter space

From: Efficient design of synthetic gene circuits under cell-to-cell variability

Criterion

MCMC

SLMCMC

Samples

20,000

200,000

Covariance

Scalar

Scalar

Diagonal

κ

–

3

{3,5}

{3,7}

{3,9}

{3,7}

Run-time (s)

23,755

900

2639

3624

3713

3413

Checking-time (s)

0

615

721

719

725

746

Rejected samples (%)

0

4

1.33

1.33

0.5

2.33

minESS

157

1505

3011

3345

2981

1487

Time/sample (s)

1.188

0.0076

0.017

0.020

0.022

0.021

Time/minESS (s)

151.31

1.01

1.12

1.19

1.49

2.80

Sampling efficiency (%)

0.79

0.75

1.51

1.67

1.49

0.74

  1. Runs were performed on a standard laptop with Intel i7 processor for the population sampling problem with \(\varepsilon = 6\textrm{nM}\). Run-time is the sampling time. Checking-time is the time to naively compute population costs for 600 sampled populations (unnecessary for the naive approach). Populations with cost above the threshold account for the rejected samples. minESS is the minimum of the individual ESSs obtained across multiple dimensions. Sampling efficiency is the ratio between minESS and the actual number of samples drawn. Ratios involving time are computed by summing over run-time and checking-time
Back to article page

BMC Bioinformatics

ISSN: 1471-2105

Contact us