# Team:SUSTC-Shenzhen-B/achievements

(Difference between revisions)
 Revision as of 03:26, 27 October 2012 (view source)Water (Talk | contribs)← Older edit Revision as of 03:54, 27 October 2012 (view source)Water (Talk | contribs) Newer edit → Line 466: Line 466:

Results

Results

-

1, Theoretical model of predicting efficiency of terminators

+

1. Theoretical model of predicting efficiency of terminators

The goal of theoretical model to predict the efficiency of a terminator from its

The goal of theoretical model to predict the efficiency of a terminator from its Line 476: Line 476:

To get a relationship between d score and efficiency, we use the biofab data as

To get a relationship between d score and efficiency, we use the biofab data as training dataset. Biofab is the International Open Facility Advancing Biotechnology, which was founded in 2009. It's led by bioengineers from UC Berkeley and Stanford. In biofab data, there are 40 terminators with efficiency data available. We calculated the d score of terminators in Biofab and plot it with efficiency (Figure 1).

training dataset. Biofab is the International Open Facility Advancing Biotechnology, which was founded in 2009. It's led by bioengineers from UC Berkeley and Stanford. In biofab data, there are 40 terminators with efficiency data available. We calculated the d score of terminators in Biofab and plot it with efficiency (Figure 1).

-
+
(Figure 1 diagram showing the correlation between d score and efficiency) (Figure 1 diagram showing the correlation between d score and efficiency) -

We found that d score and efficiency is positively correlated. Linear fitting is +

We found that d score and efficiency is positively correlated. Linear fitting by parts is performed. - performed and the fitting curve is
+ We calculate the average error as follows:
- E=1.4D+39.1   (2)
+     (2)
- The R^2 is 0.54.

+ Here Δe is the average error, n is the number of data , E（prediction） is the efficiency predicted by our model and  E（measurement） is the efficiency measured in experiment.
+ For Figure 1, the average error is 16%. + +

We proposed that terminators can be classified into three classes based on efficiency.

We proposed that terminators can be classified into three classes based on efficiency.
-
+
(Table.1 Table of  efficiency and d score correspond to different class of terminators). (Table.1 Table of  efficiency and d score correspond to different class of terminators).

2, Validating theoretical model by existing data

2, Validating theoretical model by existing data

-

We collect existing data of efficiency of terminators from literatures and databases except Biofab. Using these data, we plot the predicted efficiency based on formula (2) with efficiency data (Figure 2). We found that our theoretical predication fit experiemental data quite well with correlation coefficient 0.8
+

We collect existing data of efficiency of terminators from literatures and databases except Biofab. Using these data, we plot the predicted efficiency based on formula (2) with efficiency data (Figure 2). We found that our theoretical predication fit experiemental data quite well.
-
+
- (Figure 2, diagram of validation using our experimental data)

+ (Figure 2, diagram of validation using our experimental data)

- +

For Figure 2, we also calculate the average error. using equation(2).  The result is 18%.

3, TTEC software and webtool

3, TTEC software and webtool

We developed TTEC software to predict terminator efficiency, which is available at

We developed TTEC software to predict terminator efficiency, which is available at

Title

# Results

### 1. Theoretical model of predicting efficiency of terminators

The goal of theoretical model to predict the efficiency of a terminator from its nucleotide sequence. We use free energy methods to predict the secondary structure of terminator. A d score is defined as
D=-96.6DΔG/L+18.6T-116.9 (2)
Here, ΔG is the Gibbs free energy change of stemloop formation. L is the length of stemloop. T is the score of T tail.

To get a relationship between d score and efficiency, we use the biofab data as training dataset. Biofab is the International Open Facility Advancing Biotechnology, which was founded in 2009. It's led by bioengineers from UC Berkeley and Stanford. In biofab data, there are 40 terminators with efficiency data available. We calculated the d score of terminators in Biofab and plot it with efficiency (Figure 1). (Figure 1 diagram showing the correlation between d score and efficiency)

We found that d score and efficiency is positively correlated. Linear fitting by parts is performed. We calculate the average error as follows: (2)
Here Δe is the average error, n is the number of data , E（prediction） is the efficiency predicted by our model and E（measurement） is the efficiency measured in experiment.
For Figure 1, the average error is 16%.

We proposed that terminators can be classified into three classes based on efficiency. (Table.1 Table of efficiency and d score correspond to different class of terminators).

### 2, Validating theoretical model by existing data

We collect existing data of efficiency of terminators from literatures and databases except Biofab. Using these data, we plot the predicted efficiency based on formula (2) with efficiency data (Figure 2). We found that our theoretical predication fit experiemental data quite well. (Figure 2, diagram of validation using our experimental data)

For Figure 2, we also calculate the average error. using equation(2). The result is 18%.

### 3, TTEC software and webtool

We developed TTEC software to predict terminator efficiency, which is available at www.terminator efficiency.com (for our web tool)
https://github.com/igemsoftware/SUSTC_Shenzhen_B_2012 (for our software) (Figure 3 the interface of our web tool) (Figure 4 the intrface of outr software)

### 4, Database of terminator efficiency

We built a database of terminators whose efficiency had been measured. Not only terminator efficiency , other information such as the direction, category and structure are also included. Some of the data were from partsregistry and Biofab.Terminator Library. Others were from papers. We read many papers to find those terminators and some of the information were got through our e-mail with the author.The database will continue to update to give it more utility. （Figure 5 our database of terminators whose efficiency have been measured）