5 Models Provided in Agronomy Method

1.          A Model (G+GE)

Additive model partitions genetic variance into additive and additive × environment variance components.

2.          AD Model (A+D+AE+DE)

AdditiveDominance model, partitions genetic variance into additive, dominance variance components, as well as their interactions with environments, additive × environment, dominant × environment variance components.

3.        AD+AA Model (A+D+AA+AE+DEAAE)

AdditiveDominanceEpistasis model, partitions genetic variance into additive, dominance, and additive-additive epistasis variance components, as well as their interactions with environments, additive × environment, dominant × environment, epistasis × environment variance components.

4.        ADM Model (A+D+M+AE+DE+ME)

Additive + Dominance + Maternal model, partitions genetic variance into additive, dominance, and maternal variance components, as well as their interactions with environments, additive × environment, dominance × environment, maternal × environment variance components.

5.        ADMP Model (A+D+M+P+AE+DE+ME+PE)

Additive + Dominance + Maternal + Paternal model, partitions genetic variance into additive, dominance, maternal and paternal variance components, as well as their interactions with environments, additive × environment, dominance × environment, maternal × environment and paternal × environment variance components.

6.        AMC Model (A+M+C)

Anther Additive + Maternal + Cytoplasmic model, partitions genetic variance into anther additive (gematic additive), maternal and cytoplasmic variance components.

 

Format of input data

The format of data is shown in files CotData.txt, CotF1F2.txt, CotF2.txt in the Sample folder, and CotBoll.txt is a good example data with time-specific records for conditional analysis.

The first five fixed columns in the above data files represent environment (e.g. year or location), female, male, cross and block. From the sixth column on, trait(s) values could be listed, no more than 100 traits could be accepted in a file.

The cross code is 0 for parent, 1 for F1, 2 for F2, 3 for F1 × Pi, 4 for F1 × Pj, 5 for Pi × F1, and 6 for Pj × F1. The minimum value is 1 for all other four fixed column. Correspondingly, the maximum value for each fixed column is the total number of environments for the first column, of parents for the second and third columns, and number of blocks within each environment for the fifth column. The code should be continuous integers and arranged in order. Each missing trait value is denoted by a dot(".").

 

All the models within Agronomy method share the same Coefficient-setting Box, which will pop out automatically when you select an agronomy model from Agronomy menu.

 

1.    Have block effect

For an input file, if the block column contains more than 1 block, both No and Yes options enable. Check No for no or ignoring block effect for raw data, then the "block" will be considered as replication if more than one block included. Check Yes for involving block effect in raw data. If there is only 1 block in the raw data, the option should be No.

2.     Jackknife Kind

If there are no block effect in data or block considered as replication, Cell will be the default value of Jackknife kind. When there is block effect in data, either Block or Cell is acceptable.

3.     Jackknife Number

If the Jackknife Kind is Block, Jackknife Number resampling unit is default 1. If Jackknife Kind is Cell, Jackknife Number of resampling unit can range form 1 to 9 (1 is recommended, but for big data it will be time exhausting).

4.     Random effects predict method

AUP(adjusted unbiased prediction) or LUP(linear unbiased prediction) can be used in predicting random effects, such as additive effect, dominant effect, et al.

5.     Language

Select the Language of the output file. Both Chinese and English are available.

6.    Condition or Not

If you do not want to invoke this function, please choose No, which is default. If you have chosen the Yes option, you should better select the Condition Type value and Step number. If you choose contribution, condition type will automatically be Final, step number be 1, and only variance module will be run. No matter Yes or Contribution option checked, there should be more than 1 traits listed in an input file if you really want to employ Condition analysis.

7.     Condition Type

Step and Final are two kinds of strategies for conditional analysis. e.g., if you choose Step with step number 2 and there are four traits listed, the program will take the conditional analysis twice, calculating the trait3|trait1 (trait3 condition on trait1), trait4|trait2. If you choose Final type, the step number will equal the default 1, and if there are N traits listed in the input file, the program will sequentially calculate traitN|trait1, traitN|trait2, traitN|trait3....traitN|traitN-1.

8.    Step

Step number ranges from 1 to 9, which should be smaller than the total trait number accomodated in an input file. The certain value selected according to your input data and research interests.

9.    Var

If you check this option, the program will estimate variance components and predict random effects. The results are automatically saved in file filename.var (filename is the name of input file).

10.    Het

If you tick this option, the program will predict heterosis. The results are automatically saved in file filename.pre. For  A, ADM, ADMP, and ADGE models, this option is unavailable.

11.    Cov

If you tick this option, the program will estimate covariance components and correlation coefficients, but you should choose Var and/or Het option first. The results are automatically saved in file filename.cov.

12.    Run

Run the data with selected model and model.

13.    Cancel

Leave the Coefficient-setting Box, and do nothing.

 

 

 

Home Next