Team:USTC-Software/Project/Examples

From 2013.igem.org

Revision as of 10:44, 27 September 2013 by Zigzag (Talk | contribs)

Slide

Take a gNAP before wearing your gloves! Genetic Network Analyze and Predict
The sketch and final GUI of gNAP!
We compare the result of our software with gene expression profile in literature.
We are USTC-Software!

Example

Examples

To prove our software’s reliability, we search for literatures to test our program. It is hard to find an appropriate literature which studies the effect of importing an exogenous gene into E.coli K-12. But actually, our software could also simulate the effect of changing endogenous gene by putting the same promoter and gene sequence in.

In the literature we found, Stuart and his team measured the gene expression profiles in otherwise isogenic integration host factor IHF+ and IHF- strains. And IHF is one of the genes in our genetic regulatory network(GRN).

By importing the IHF’s promoter and gene sequence, we used our software simulating the enhancement of IHF’s expression and compared the result with the gene expression profile in that literature.

There are 30 genes in that profile which are also in our GRN. Here is the list and Genes differentially expressed between E. coli K12 strains IH100 (IHF+) and IH105 (IHF-) with a p value less than 0.0005:

Gene

Avg

S.D.

p value

Fold

Compare
Result

IH100

IH105

IH100

IH105

glnA

2.91E-03

9.39E-04

6.80E-04

1.33E-04

1.30E-03

-3.1

fit

ilvA

5.06E-04

3.42E-04

1.86E-05

2.26E-05

3.00E-05

-1.48

unfit

ilvE

5.81E-04

3.58E-04

4.70E-05

5.77E-05

9.80E-04

-1.62

unfit

ilvG

1.97E-04

7.67E-05

2.65E-05

2.23E-05

4.40E-04

-2.57

unfit

leuA

6.99E-04

1.07E-03

9.21E-05

9.23E-05

1.30E-03

1.53

fit

cobT

1.00E-05

7.97E-05

7.82E-06

2.13E-05

8.50E-04

7.97

unfit

cobU

4.26E-05

1.22E-04

1.79E-05

1.95E-05

9.90E-04

2.85

unfit

lacA

5.14E-03

1.21E-03

1.54E-03

3.52E-04

2.50E-03

-4.24

unfit

lacZ

2.10E-03

5.14E-04

3.77E-04

1.34E-04

2.20E-04

-4.08

unfit

lacY

1.62E-03

4.08E-04

2.53E-04

7.95E-05

9.80E-05

-3.96

unfit

ompF

7.23E-03

2.35E-03

1.90E-03

3.69E-04

2.40E-03

-3.07

fit

gltD

9.91E-04

1.40E-04

1.88E-04

3.06E-05

1.10E-04

-7.1

fit

lpdA

1.07E-03

7.60E-04

1.17E-04

7.75E-05

4.60E-03

-1.41

fit

rffT

5.81E-06

3.65E-05

4.66E-06

2.86E-05

9.40E-04

6.28

fit

ndh

5.03E-05

1.46E-04

1.94E-05

3.29E-05

2.50E-03

2.9

fit

cheR

1.29E-04

2.68E-05

2.07E-04

1.75E-05

1.30E-03

-4.82

fit

sodA

3.80E-04

9.74E-04

1.06E-04

6.26E-05

7.00E-05

2.57

fit

sodB

7.80E-04

1.91E-03

2.45E-04

4.11E-04

3.30E-03

2.44

fit

cpdB

1.92E-05

7.56E-05

1.24E-05

1.40E-05

9.50E-04

3.94

fit

guaA

8.25E-04

4.31E-04

5.43E-05

1.34E204

1.60E203

-1.91

unfit

yiaJ

3.47E-05

6.15E-04

1.74E-05

1.64E204

4.10E204

17.74

fit

dsdX

1.05E-05

3.88E-05

5.23E-06

2.44E205

1.70E203

3.7

fit

oppD

2.32E-05

8.02E-05

1.81E-05

1.66E205

3.50E203

3.46

fit

glnL

2.41E-04

3.99E-05

4.81E-05

2.81E205

3.60E204

-6.04

fit

oppA

2.54E-03

5.06E-03

1.72E-04

5.68E204

1.40E204

2

fit

oppB

1.06E-04

3.57E-04

3.06E-05

6.22E205

3.60E204

3.35

fit

proV

2.50E-05

5.30E-05

7.34E-06

9.57E206

3.60E203

2.12

fit

rbsC

4.20E-05

1.12E-04

1.47E-05

2.70E205

3.90E203

2.67

fit

hdeB

1.09E-03

5.51E-06

1.80E-04

3.47E206

2.00E205

-198.5

fit

yefM

4.63E-04

8.12E-04

5.02E-05

6.07E205

1.10E204

1.75

fit

The comparing result means that whether the result of our software fits to the result of gene expression profile. After statistic, in these 30 genes, there are 21 genes whose result are same to gNAP's simulation, 70% of the total.
What’s more, it is easy to see that the result unfitted often from the same series of genes, such as ilv, cob, lac. After integrating those genes, the degree of fitness increased to 84%.
Therefore, we may draw the following conclusion that our software could simulate the impact of new gene to some extent.

Consistency

The consistency of the program has also been tested. We inserted a gene as same as a gene in the network and compared the regulations predicted by the program with the original regulations. Without filtering the random similarities, the actual regulations were submerged in the network noise.

Figure 1. The green line represents regulating values.
The blue line represents regulated values.

Figure 2.Predicted regulation without filtered.
The actual regulations are submerged by the noise.

With random similarities filtered, all original regulations were picked out. The result shows that the program is consistent with the original network.


Figure 3.The SNR is better.
The actual regulations are picked out.

Reference

Arfin S M, Long A D, Ito E T, et al. Global Gene Expression Profiling in Escherichia coliK12 THE EFFECTS OF INTEGRATION HOST FACTOR[J]. Journal of Biological Chemistry, 2000, 275(38): 29672-29684.