Home

User`s Manual

1. 21 e The variance components and broad sense heritability of the MRV by location will be stored in Heritability Yield for Testing and Deleting Locations Option 41 csv e The phenotypic correlations among locations for the MRV will be stored in Phen Corr for Yield Among Locations Option 41 csv e The genetic correlations among locations for the MRV will be stored in Genetic Correlations for Yield Among Locations Option 41 csv e A dendrogram of the cluster analysis will be stored in Tree Yield RCBD Option 41 CGM e The plot of the first 2 principal components will save to PCA Yield RCBD Option 41 CGM Option 42 BLUEs amp BLUPs Individual Analysis without Management RCBD without covariate This option will calculate BLUEs and BLUPs for the MRV without adjusting for a covariate BLUEs broad sense heritability LSD grand mean variance components and coefficients of variation for all of your variables It will not separate out different management methods To run it type 4 and then 42 from the Data visualization vs analysis menu OUTPUT TO SCREEN e BLUEs and BLUPs for the MRV without adjusting for a covariate BLUEs broad sense heritability LSD grand mean variance components and CV for all variables in each location OUTPUT TO FILE e Asummary file with BLUEs and BLUPs for the MRV without adjusting for a covariate BLUEs broad sense heritability LSD grand mean variance components
2. Option 1 Lattice Analysis with Covariate Adjustment 11 Option 11 Genetic Correlations Among Locations Lattice 11 Option 12 BLUEs BLUPs Individual Analysis without Management Lattice 12 Option 13 BLUEs BLUPs Individual Analysis with Management Lattice 12 Option 14 BLUEs BLUPs Combined Analysis By Management Lattice 12 Option 15 BLUEs BLUPs Combined Analysis Across All Locations Lattice 13 Option 16 Running All Options in a Simple Step Lattice 13 Option 17 Exit META uu a ata as 14 Option 2 Randomized Complete Block Design with Covariate Adjustment 14 Option 21 Genetic Correlations Among Locations RCBD 14 Option 22 BLUEs BLUPs Individual Analysis without Management RCBD 15 Option 23 BLUEs BLUPs Individual Analysis with Management RCBD 15 Option 24 BLUEs BLUPs Combined Analysis By Management RCBD 15 Option 25 BLUEs BLUPs C
3. analysis menu 14 Option 17 Exit META This option will exit META Option 2 Randomized Complete Block Design with Covariate Adjustment Option 21 Genetic Correlations Among Locations RCBD This option will calculate the genetic correlations between locations for the MRV adjusted by one covariate that you specify in the data input step It will also perform a cluster and PCA analysis based on the distance matrix To run it type 2 and then 21 from the Data visualization vs analysis menu CHANGES TO SAS FILE e Change file Option 21 Genetic Correlations Among Locations RCBD sas e By default the dendrogram and PCA biplot are printed to a powerpoint editable CGM file If you wish for them to print to the screen instead o add a sign to lines 428 429 and 430 dendrogram and to lines 525 526 and 527 PCA biplot o change the length of axis2 line 534 to 3 5 in from 7 0 in o change the length of axis1 line 542 to 8 0 in from 9 0 in OUTPUT TO SCREEN e Variance components and broad sense heritability by location o Note if CodeForOut is 999 this means that the location s heritability was below your cutoff threshold and has been deleted from the analysis e ANOVA to verify the final locations retained in the data e Phenotypic correlations for the MRV with p values e Phenotypic correlations for the MRV without p values e Genetic correlations for the MRV among locations e Eigenvalues for cluster analysis e Ei
4. ee ren Zo eee T ustsaq 291187 usiseq l duroS pezmopuey pas upISoid uni SNU9A VLAN youney Notes Once you finish running one of the four alternative analyses depicted in the above flow diagram Lattice or RCBD with or without covariate if you want to run other analyses it is necessary first to close the current SAS session and start again from the beginning This is done in order to clean all the SAS work memory between analyses No matter which option you run the program will always start by printing descriptive statistics univariate statistics for each variable in the data set univariate statistics for each variable kept in the data set and the records for the first 25 entries use these to verify that the data have been entered correctly As the option runs SAS will print out covariance parameter estimates and type 3 tests of fixed effects each time a mixed model is run there are titles above each to explain which test was performed When covariance parameters grand mean LSD CV and broad sense heritability are calculated they are only calculated once for your MRV Therefore the columns under BLUP_MRV will be blank When data are combined across locations any locations with a heritability of less than 0 05 will be eliminated If you wish to use a different cutoff you must change to SAS program 10 lattice with covariate 20 RCBD with covariate 30 lattice no covariate or 40 RCDB no covariate depend
5. will be stored in BLUEs amp BLUPs Individual Analysis without Management Option 22 csv Option 23 BLUEs amp BLUPs Individual Analysis with Management RCBD This option will calculate BLUEs and BLUPs for the MRV BLUEs broad sense heritability LSD grand mean variance components and coefficients of variation for all of your variables it will correct the MRV for your covariate It gives the same results as option 22 but labels the locations by management type To run it type 2 and then 23 from the Data visualization vs analysis menu OUTPUT TO SCREEN e Grand means by location and management for the MRV after adjusting for the covariate in each location e BLUEs and BLUPs for the MRV BLUEs broad sense heritability LSD grand mean variance components and CV for all variables in each location OUTPUT TO FILE e Asummary file with BLUEs and BLUPs for the MRV BLUEs broad sense heritability LSD grand mean variance components and CV for all variables in each location will be stored in BLUEs amp BLUPs Individual Analysis with Management Option 23 csv Option 24 BLUEs amp BLUPs Combined Analysis By Management RCBD This option will calculate BLUEs and BLUPs for the MRV BLUEs broad sense heritability LSD grand mean variance components and coefficients of variation for all of your variables it will correct the MRV for your covariate It will not break down the data by location but will combine them acro
6. Option 31 csv e The phenotypic correlations among locations for the MRV will be stored in Phen Corr for Yield Among Locations Option 31 csv e The genetic correlations among locations for the MRV will be stored in Genetic Correlations for Yield Among Locations Option 31 csv e A dendrogram of the cluster analysis will be stored in Tree Yield Lattice Option 31 CGM e The plot of the first 2 principal components will save to PCA Yield Lattice Option 31 CGM 18 Option 32 BLUEs amp BLUPs Individual Analysis without Management Lattice without covariate This option will calculate BLUEs and BLUPs for the MRV without adjusting for a covariate BLUEs broad sense heritability LSD grand mean variance components and coefficients of variation for all of your variables It will not separate out different management methods To run it type 3 and then 32 from the Data visualization vs analysis menu OUTPUT TO SCREEN e BLUEs and BLUPs for the MRV without adjusting for a covariate BLUEs broad sense heritability LSD grand mean variance components and CV for all variables in each location OUTPUT TO FILE e Asummary file with BLUEs and BLUPs for the MRV without adjusting for a covariate BLUES broad sense heritability LSD grand mean variance components and CV for all variables in each location will be stored in BLUEs BLUPs Individual Analysis without Management Option 32 csv Option 33 B
7. broad sense heritability LSD grand mean variance components and coefficients of variation for all of your variables It will not break down the data by management condition but will combine across all locations If you do not have different management conditions use this option To run it type 4 and then 45 from the Data visualization vs analysis menu OUTPUT TO SCREEN e Variance components and broad sense heritability for the MRV without adjusting for a covariate by location this will tell you if any of your locations is being suppressed due to poor heritability e BLUEs and BLUPs for the MRV without adjusting for a covariate and BLUEs for all other variables Nreps nlocs location variance genotype variance locationxgenotype variance 23 residual variance grand mean LSD CV and broad sense heritability for each trait combined across all locations OUTPUT TO FILE e A file with variance components and broad sense heritability estimates by location for the MRV will be stored in Heritability Yield for Testing and Deleting Locations Option 45 csv e A file with BLUEs and BLUPs for the MRV without adjusting for a covariate and BLUEs for all other variables Nreps nlocs location variance genotype variance location genotype variance residual variance grand mean LSD CV and broad sense heritability for each trait combined across all locations will be stored in BLUEs amp BLUPs Combined Analysis Across All Loc
8. LSD grand mean variance components and CV for all variables in each location will be stored in BLUEs amp BLUPs Individual Analysis without Management Option 12 csv Option 13 BLUEs amp BLUPs Individual Analysis with Management Lattice This option will calculate BLUEs and BLUPs for the MRV BLUEs broad sense heritability LSD grand mean variance components and coefficients of variation for all of your variables it will correct the MRV for your covariate It gives the same results as option 12 but labels the locations by management type To run it type 1 and then 13 from the Data visualization vs analysis menu OUTPUT TO SCREEN e Grand means by location and management for the MRV after adjusting for the covariate in each location e BLUEs and BLUPs for the MRV BLUEs broad sense heritability LSD grand mean variance components and CV for all variables in each location OUTPUT TO FILE e Asummary file with BLUEs and BLUPs for the MRV BLUEs broad sense heritability LSD grand mean variance components and CV for all variables in each location will be stored in BLUEs amp BLUPs Individual Analysis with Management Option 13 csv Option 14 BLUEs amp BLUPs Combined Analysis By Management Lattice This option will calculate BLUEs and BLUPs for the MRV BLUEs broad sense heritability LSD grand mean variance components and coefficients of variation for all of your variables it will correct the MRV for you
9. LUEs amp BLUPs Individual Analysis with Management Lattice without covariate This option will calculate BLUEs and BLUPs for the MRV without adjusting for a covariate BLUEs broad sense heritability LSD grand mean variance components and coefficients of variation for all of your variables It gives the same results as option 32 but labels the locations by management type To run it type 3 and then 33 from the Data visualization vs analysis menu OUTPUT TO SCREEN e Grand means by location and management for the MRV without adjusting for a covariate e BLUEs and BLUPs for the MRV without adjusting for a covariate BLUEs broad sense heritability LSD grand mean variance components and CV for all variables in each location OUTPUT TO FILE e Asummary file with BLUEs and BLUPs for the MRV without adjusting for a covariate BLUEs broad sense heritability LSD grand mean variance components and CV for all variables in each location will be stored in BLUEs amp BLUPs Individual Analysis with Management Option 33 csv Option 34 BLUEs amp BLUPs Combined Analysis By Management Lattice without covariate This option will calculate BLUEs and BLUPs for the MRV without adjusting for a covariate BLUEs broad sense heritability LSD grand mean variance components and coefficients of variation for all of your 19 variables It will not break down the data by location but will combine them across management conditions To
10. User s Manual META Multi Environmental Trial Analysis of Plant Breeding Data Mateo Vargas Emily Combs Gregorio Alvarado Gary Atlin and Jose Crossa This set of SAS options will generate graphics BLUPs BLUEs LSDs CVs broad sense heritability and other statistics for breeding trials that are run in an RCBD or a Lattice design The main response variable MRV may be adjusted by one or more covariates Analyses may be done by location across management conditions or across all locations Once your data entry file is prepared based on the directions below the entire program is run from a graphical user interface These are basic instructions more detailed documentation is available in the SAS code itself Table of Contents GELLING Started miii iii 4 NOTES ucrrcordara assaka naaswanaypanshasqauqaaayayqaaussuysa Qasqkaesyakapauyasaqykasiuyaataqakqa ayauya Rasuka OEE E EEA tiyaya sa Eo esien sis 6 Preparing your data tor METAL u uqa a a aaa qusa ca 7 RUNNING the Pro Mireia da 8 Data Visualiz titi la ista ae iaaa Dia dear il iva 10 Option 01 and 03 Box PIOtS cccccccccsssssssececeesesssssaesececessussaeseseceesessussaeececessesesasaceeceeessessanaeseceeeesees 10 Option 02 and 04 Frequency Histograms 10 Option 00 Back to Previous M NuU cscssscccecesssssssaeeeceescesseeeeeececeseesaeaeaeeeescesesaeaeeeeeessseseneaeeeseesees 10 DA e 11
11. V for your covariate It will not break down the data by management conditions but will combine across all locations If you do not have different management conditions use this option To run it type 1 and then 15 from the Data visualization vs analysis menu OUTPUT TO SCREEN e Variance components and broad sense heritability for the MRV by location this will tell you if any of your locations is being suppressed due to poor heritability e BLUEs and BLUPs for the MRV and BLUEs for all other variables Nreps nlocs location variance genotype variance locationxgenotype variance residual variance grand mean LSD CV and broad sense heritability for each trait combined across all locations OUTPUT TO FILE e A file with variance components and broad sense heritability estimates by location for the MRV will be stored in Heritability Yield for Testing and Deleting Locations Option 15 Lattice CSV e A file with BLUEs and BLUPs for the MRV and BLUEs for all other variables Nreps nlocs location variance genotype variance locationxgenotype variance residual variance grand mean LSD CV and broad sense heritability for each trait combined across all locations will be stored in BLUEs amp BLUPs Combined Analysis Across All Locations Option 15 Lattice CSV Option 16 Running All Options in a Simple Step Lattice This option will run all of the above options To run it type 1 and then 16 from the Data visualization vs
12. and CV for all variables in each location will be stored in BLUEs amp BLUPs Individual Analysis without Management Option 42 csv Option 43 BLUEs amp BLUPs Individual Analysis with Management RCBD without covariate This option will calculate BLUEs and BLUPs for the MRV without adjusting for a covariate BLUEs broad sense heritability LSD grand mean variance components and coefficients of variation for all of your variables It gives the same results as option 42 but labels the locations by management type To run it type 4 and then 43 from the Data visualization vs analysis menu OUTPUT TO SCREEN e Grand means by location and management for the MRV without adjusting for a covariate e BLUEs and BLUPs for the MRV without adjusting for a covariate BLUEs broad sense heritability LSD grand mean variance components and CV for all variables in each location OUTPUT TO FILE 22 e Asummary file with BLUEs and BLUPs for the MRV without adjusting for a covariate BLUEs broad sense heritability LSD grand mean variance components and CV for all variables in each location will be stored in BLUEs amp BLUPs Individual Analysis with Management Option 43 csv Option 44 BLUEs amp BLUPs Combined Analysis By Management RCBD without covariate This option will calculate BLUEs and BLUPs for the MRV without adjusting for a covariate BLUEs broad sense heritability LSD grand mean variance components and coefficie
13. ations Option 45 csv Option 46 Running All Options in a Simple Step RCBD without covariate This option will run all options that start with a 4 To run it type 4 and then 46 from the Data visualization vs analysis menu Option 47 Exit META This option will exit META 24 Example using Lattice Dataset What follows are screenshots showing how to run the sample lattice dataset and get a complete data analysis with yield as the MRV Yield will also be adjusted by anthesis date Open META Menus and hit run SAS File Edit View Tools Run Solutions Help YA Contents of SAS Environment g G Libraries File Shortcuts Favorite Folders wawawawnwawawaasanaqaanasawwanwaaqaanasawaqawaa AAN AAA ANNAN Ay Praia Samira wari raros HIE KOR an THEN TOLEDO THE ON SCREEN WENT pi Computer RRR RRR RRR AAA ARA AAA AEREA Ad REAR ARERR RRA RRA AAA ARA ARA ARA RARA RARA RARA RRA RRA ERRE options ps 500 ls 120 nodate nonumber Menu for Selection of Experimental Design Output Untitled E Log Untitled Editor Untitled META Menus E CASAS PROGRAMS GMP Ln 5 Col 19 The Select Design menu will pop up am using a lattice design so selected option 1 EXPERIMENTAL DESIGN OPTIONS Write 1 for LATTICE Write 2 for RANDOMIZED COMPLETE BLOCKS OPTION SELECTED 1__ would like to adjust by a co
14. break down the data by management condition but will combine them across all locations If you do not have different management conditions use this option To run it type 3 and then 35 from the Data visualization vs analysis menu OUTPUT TO SCREEN e Variance components and broad sense heritability for the MRV by location this will tell you if any of your locations is being suppressed due to poor heritability e BLUEs and BLUPs for the MRV without adjusting for a covariate and BLUEs for all other variables Nreps nlocs location variance genotype variance locationxgenotype variance residual variance grand mean LSD CV and broad sense heritability for each trait combined across all locations OUTPUT TO FILE e A file with variance components and broad sense heritability estimates by location for the MRV will be stored in Heritability Yield for Testing and Deleting Locations Option 35 csv e A file with BLUEs and BLUPs for the MRV without adjusting for a covariate and BLUEs for all other variables Nreps nlocs location variance genotype variance locationxgenotype variance 20 residual variance grand mean LSD CV and broad sense heritability for each trait combined across all locations will be stored in BLUEs amp BLUPs Combined Analysis Across All Locations Option 35 csv Option 36 Running All Options in a Simple Step Lattice without covariate This option will run every option that starts with a 3 To ru
15. do this step very carefully as any errors will result in program failure You can correct any field before pressing ENTER in the last field of the menu i e in the Genotype field You can move through the fields using the mouse or the arrow keys from the keyboard a Enter the location of your options folder this is where you saved all the SAS options e g C Users John Doe Awesome SAS Options b Enter the name of your data file The data file must be a csv file but you should not include csv e g if your file is called My Great Data csv type in My Great Data c Enter the name of the folder where you want the results to be saved The folder must exist before running the programs because SAS cannot create it e g C Users John Doe Awesome SAS Options Results d Depending on the design you are running lattice vs randomized complete block you will need to tell META what you called the following factors and variables For example in your data entry file you may have called the column containing information on location Locat or L or Locations e Main Response Variable MRV e Covariate if using a covariate e Grouping factor if using a grouping factor e Location e Replication e Block e Genotype e Press enter or carriage return after you have entered each variable name or move to the next field using the mouse or arrow keys META will assume any other columns in your data s
16. e of the column containing location assignments Replication The name of the column containing replication assignments Block The name of the column containing block assignments Entry The name of the column containing genotype names Other variables META will calculate BLUEs for any other variable e g flowering dates kernel composition traits at NOTE Do not use names Loc Rep Iblock or Genotype to identify u the factors Location Replicate Block Genotype in your csv file as they are used as reserve words in the META programs Running the program Open META Menus in SAS a Hit run and follow the on screen menu b More detailed instructions for each option are available below and in comments in the code For standard use there is no need to modify any of the SAS codes Advanced users may wish to do so based on instructions provided below or comments in the code c Inthe on screen menu type the requested information into the white line next to the sign You may need to click on the white line in order to type text there Select Design menu a Fora lattice design type 1 b Fora randomized complete blocks design type 2 Select Covariate menu a Tousea covariate type 1 b To analyze without a covariate type 2 Data Entry menu tell META where your files are located and what you named your factors and variables Please
17. et contain additional trait phenotypes and will calculate BLUEs and other statistics Note You can correct any field before pressing the Genotype field After this if the program finds any inconsistencies among the names given in the fields of the last menu and the name given in the input path output path input file main response variable covariate and factors in the csv data file it will send an ERROR message indicating what was wrong Pressing ENTER makes the program go back to the last menu that can be corrected and continue with the analyses The ERROR message will appear only once each time so that if you have several mistakes you can verify and correct them one by one If you detect several mistakes in the same step you can correct all of them at the same time e Data visualization vs analysis menu a b To see options for data visualization type 0 To see options for data analysis type the suggested number 1 2 3 or 4 depending on the experiment design type with or without covariate Note if you wish to both visualize and analyze your data visualization must be done first The analysis sub menu will not allow you to return to the main menu Details of visualization and analysis options are listed below The data visualization step is the same for all four analysis types lattice and RCBD designs with and without covariates The data analysis steps are different depending on the experimental design 10 Data V
18. genvalues for the PCA OUTPUT TO FILE e The variance components and broad sense heritability of the MRV by location will be stored in Heritability Yield for Testing and Deleting Locations Option 21 csv e The phenotypic correlations among locations for the MRV will be stored in Phen Corr for Yield Among Locations option 21 csv e The genetic correlations among locations for the MRV will be stored in Genetic Correlations for Yield Among Locations Option 21 csv 15 e Adendrogram of the cluster analysis will be stored in Tree Yield RCBD Option 21 CGM e The plot of the first 2 principal components will save to PCA Yield RCBD Option 21 CGM Option 22 BLUEs amp BLUPs Individual Analysis without Management RCBD This option will calculate BLUEs and BLUPs for the MRV BLUEs broad sense heritability LSD grand mean variance components and coefficients of variation for all of your variables it will correct the MRV for your covariate It will not separate out different management methods To run it type 2 and then 22 from the Data visualization vs analysis menu OUTPUT TO SCREEN e BLUEs and BLUPs for the MRV BLUEs broad sense heritability LSD grand mean variance components and CV for all variables in each location OUTPUT TO FILE e Asummary file with BLUEs and BLUPs for the MRV BLUEs broad sense heritability LSD grand mean variance components and CV for all variables in each location
19. ice Analysis without Covariate Adjustment Option 31 Genetic Correlations Among Locations Lattice without covariate This option will calculate the genetic correlations between locations for the MRV without adjusting for a covariate It will also perform a cluster and PCA analysis based on the distance matrix To run it type 3 and then 31 from the data visualization vs analysis menu CHANGES TO SAS FILE e By default the dendrogram and PCA biplot are printed to a powerpoint editable CGM file If you wish for them to print to the screen instead o add a sign to lines 415 416 and 417 dendrogram and to lines 512 513 and 514 PCA biplot change the length of axis2 line 521 to 3 5 in from 7 0 in change the length of axis1 line 529 to 8 0 in from 9 0 in OUTPUT TO SCREEN e Variance components and broad sense heritability by location o Note if CodeForOut is 999 this means that the location s heritability was below your cutoff threshold and has been deleted from the analysis e ANOVA to verify the final locations retained in the data e Phenotypic correlations for the MRV with p values e Phenotypic correlations for the MRV without p values e Genetic correlations for the MRV among locations e Eigenvalues for cluster analysis e Eigenvalues for the PCA OUTPUT TO FILE e The variance components and broad sense heritability of the MRV by locations will be stored in Heritability Yield for Testing and Deleting Locations
20. ill combine them across all locations If you do not have different management conditions use this option To run it type 2 and then 25 from the Data visualization vs analysis menu OUTPUT TO SCREEN e Variance components and broad sense heritability for the MRV by location this will tell you if any of your locations is being suppressed due to poor heritability e BLUEs and BLUPs for the MRV and BLUEs for all other variables Nreps nlocs location variance genotype variance locationxgenotype variance residual variance grand mean LSD CV and broad sense heritability for each trait combined across all locations OUTPUT TO FILE e A file with variance components and broad sense heritability estimates by location for the MRV will be stored in Heritability Yield for Testing and Deleting Locations Option 25 csv e A file with BLUEs and BLUPs for the MRV and BLUEs for all other variables Nreps nlocs location variance genotype variance locationxgenotype variance residual variance grand mean LSD CV and broad sense heritability for each trait combined across all locations will be stored in BLUEs amp BLUPs Combined Analysis Across All Locations Option 25 csv Option 26 Running All Options in a Simple Step Lattice This option will run all options that start with a 2 To run it type 2 and then 26 from the Data visualization vs analysis menu 17 Option 27 Exit META This option will exit META Option 3 Latt
21. ing on which design you are using To change the cutoff change the macro variable H2_ Level on line 58 Preparing your data for META e Putallof your data into a csv file please note the option will not start reading until the an row of data use the first row for headers e Replace any missing values with for both numeric and character variables this lets SAS know this datapoint is missing Blank cells NA s or any other missing value indicators will result in option failure e Make note of your variable names e g if you called the column storing location information Locat vs location vs locations as you will need them to enter your data into the option Variable names are not case sensitive e META can accept the following variables do not include any other variables e g Range and Row numbers 1 ONO E Main Response Variable MRV The name of the column containing your major response variable most commonly yield This variable will have both BLUPs and BLUEs calculated and may be adjusted by a covariate Covariate The name of the column containing your covariate Grouping Factor If you would like some of your analyses to be grouped by a factor e g Management condition country of testing year of testing list the name of the factor here If you do not wish to group your results by a factor write None exclusively no other name is allowed Location The nam
22. isualization Options 01 and 03 Box Plots This option produces boxplots for your data by variable e g days to anthesis plant height Option 01 is for lattice designs and 03 for RCBD To run it select 0 from the Data visualization vs analysis menu and then type 01 OUTPUT TO SCREEN Boxplots for each variable by location including the basic statistics as minimum mean maximum and standard deviation OUTPUT TO FILE None Options 02 and 04 Frequency Histograms This option produces univariate statistics for each of your variable location combinations with the option to separate out replications as well It also produces frequency histograms for each variable location combination using reps as a classification variable Option 02 is for lattice designs and 04 for RCBD To run it select 0 from the Data visualization vs analysis menu and then type 02 CHANGES TO SAS FILE e The default for univariate parameters mean median outliers etc is to run them by location and variable if you wish to also run it by replication move the from line 39 to line 40 OUTPUT TO SCREEN e Univariate data mean median mode high and low values etc separated by location and variable e Frequency histograms by location and variable with the replications shown side by side OUTPUT TO FILE None Option 00 Back to Previous Menu Returns you to the data visualization vs analysis menu 11 Data Analysis The following
23. ite 1 for LATTICE ANALYSIS WITH COVARIATE OPTION SELECTED _1_ Next chose option 16 to do all the analyses LATTICE WITH COVARIATE ANALYSES OPTIONS Write 11 for Genetic Correlations Among Locations Write 12 for BLUEs amp BLUPs Individual Analysis without Management Write 13 for BLUEs amp BLUPs Individual Analysis with Management Write 14 for BLUEs amp BLUPs Combined Analysis By Management Write 15 for BLUEs amp BLUPs Combined Analysis Across All Locations Write 16 for Running All Programs in a Simple Step Write 17 for Exit META OPTION SELECTED 16 The program will then output results both to the screen and to files How to use more than one covariate To use a second or third covariate you must do two things 1 Inthe data entry programs 10 20 30 or 40 you need to add to the macro statements after the statement macro H2_Level 0 05 on line 58 to declare a second covariate and state what it is In this example our second covariate is ASI macro Covar2 ASI 27 2 Anywhere that covar1 is included you would need to add covar2 as well For example option 11 on lines 37 45 would need to be modified to Titlel Analysis for Heritability proc mixed method reml data A11 Sort noinfo noitprint covtest by Loc class Rep Iblock Genotype model Yield amp Covarl HPA random Genotype Rep Iblock Rep ods output covparms COV_GY ClassLevels Levels GY ods listing excl
24. n it type 3 and then 36 from the Data visualization vs analysis menu Option 37 Exit META This option will exit META Option 4 Randomized Complete Block Design without Covariate Adjustment Option 41 Genetic Correlations Among Locations RCBD without covariate This option will calculate the genetic correlations between locations for the MRV without adjusting for a covariate It will also perform a cluster and PCA analysis based on the distance matrix To run it type 4 and then 41 from the Data visualization vs analysis menu CHANGES TO SAS FILE e By default the dendrogram and PCA biplot are printed to a powerpoint editable CGM file If you wish for them to print to the screen instead o add a sign to lines 420 421 and 422 dendrogram and to lines 517 518 and 519 PCA biplot o change the length of axis2 line 526 to 3 5 in from 7 0 in o change the length of axis1 line 534 to 8 0 in from 9 0 in OUTPUT TO SCREEN e Variance components and broad sense heritability by location o Note if CodeForOut is 999 this means that the location s heritability was below your cutoff threshold and has been deleted from the analysis e ANOVA to verify the final locations retained in the data e Phenotypic correlations for the MRV with p values e Phenotypic correlations for the MRV without p values e Genetic correlations for the MRV among locations e Eigenvalues for cluster analysis e Eigenvalues for the PCA OUTPUT TO FILE
25. nts of variation for all of your variables It will not break down the data by location but will combine them across management conditions To run it type 4 and then 44 from the Data visualization vs analysis menu OUTPUT TO SCREEN e Variance components and broad sense heritability for the MRV by locations this will tell you if any of your locations is being suppressed due to poor heritability e BLUEs and BLUPs for the MRV without adjusting for a covariate BLUES for all other variables Nreps nlocs location variance genotype variance locationxgenotype variance residual variance grand mean LSD CV and broad sense heritability for each trait management combination OUTPUT TO FILE e A file with variance components and broad sense heritability estimates by location for the MRV will be stored in Heritability Yield for Testing and Deleting Locations Option 44 csv e A file with BLUEs and BLUPs for the MRV without adjusting for a covariate and BLUEs for all other variables Nreps nlocs location variance genotype variance locationxgenotype variance residual variance grand mean LSD CV and broad sense heritability for each trait management combination will be stored in BLUEs amp BLUPs Combined Analysis By Management Option 44 csv Option 45 BLUEs amp BLUPs Combined Analysis Across All Locations RCBD without covariate This option will calculate BLUEs and BLUPs for the MRV without adjusting for a covariate BLUEs
26. ombined Analysis Across All Locations RCBD 16 Option 26 Running All Options in a Simple Step Lattice 16 OpuUn 2T EKI META mrina u Ga sapa San E yea ndean ted Sots A pss 17 Option 3 Lattice Analysis without Covariate Adjustment 17 Option 31 Genetic Correlations Among Locations Lattice without covariate 17 Option 32 BLUEs amp BLUPs Individual Analysis without Management Lattice without covariate 18 Option 33 BLUEs BLUPs Individual Analysis with Management Lattice without covariate 18 Option 34 BLUEs BLUPs Combined Analysis By Management Lattice without covariate 18 Option 35 BLUEs amp BLUPs Combined Analysis Across All Locations Lattice without covariate 19 Option 36 Running All Options in a Simple Step Lattice without covariate 20 Option 372 EXIEMETA zu u u tia he aa Tua de Weed de reek ae a i e 20 Option 4 Randomized Complete Block Design without Covariate Adjustment 20 Option 41 Genetic Correlations Among Locations RCBD without covariate 20 Option 42 BLUEs amp BLUPs Individual Analysis without Management RCBD without covariate 21 Option 43 BLUEs BLUP
27. r covariate It will not break down the data by location but will combine it across management conditions To run it type 1 and then 14 from the data visualization vs analysis menu OUTPUT TO SCREEN 13 e Variance components and broad sense heritability for the MRV by location this will tell you if any of your locations is being suppressed due to poor heritability e BLUEs and BLUPs for the MRV and BLUEs for all other variables Nreps nlocs location variance genotype variance locationxgenotype variance residual variance grandmean LSD CV and broad sense heritability for each trait management combination OUTPUT TO FILE e A file with variance components and broad sense heritability estimates by location for the MRV will be stored in Heritability Yield for Testing and Deleting Locations Option 14 csv e A file with BLUEs and BLUPs for the MRV and BLUES for all other variables Nreps nlocs location variance genotype variance locationxgenotype variance residual variance grandmean LSD CV and broad sense heritability for each trait management combination will be stored in BLUEs amp BLUPs Combined Analysis By Management Option 14 csv Option 15 BLUEs amp BLUPs Combined Analysis Across All Locations Lattice This option will calculate BLUEs and BLUPs for the MRV BLUEs broad sense heritability LSD grand mean variance components and coefficients of variation for all of your variables it will correct the MR
28. run it type 3 and then 34 from the Data visualization vs analysis menu QUTPUT T0 SCREEN e Variance components and broad sense heritability for the MRV by location this will tell you if any of your locations is being suppressed due to poor heritability e BLUEs and BLUPs for the MRV without adjusting for a covariate and BLUEs for all other variables Nreps nlocs location variance genotype variance locationxgenotype variance residual variance grand mean LSD CV and broad sense heritability for each trait management combination OUTPUT TO FILE e A file with variance components and broad sense heritability estimates by location for the MRV will be stored in Heritability Yield for Testing and Deleting Locations Option 34 csv e A file with BLUEs and BLUPs for the MRV without adjusting for a covariate and BLUEs for all other variables Nreps nlocs location variance genotype variance locationxgenotype variance residual variance grand mean LSD CV and broad sense heritability for each trait management combination will be stored in BLUEs amp BLUPs Combined Analysis By Management Option 34 csv Option 35 BLUEs amp BLUPs Combined Analysis Across All Locations Lattice without covariate This option will calculate BLUEs and BLUPs for the MRV without adjusting for a covariate BLUEs broad sense heritability LSD grand mean variance components and coefficients of variation for all of your variables It will not
29. s Individual Analysis with Management RCBD without covariate 21 Option 44 BLUEs amp BLUPs Combined Analysis By Management RCBD without covariate 22 Option 45 BLUEs BLUPs Combined Analysis Across All Locations RCBD without covariate 22 Option 46 Running All Options in a Simple Step RCBD without covariate 23 Optoma EX META oct dt Qy S RS aQ tues td al 23 Example using Lattice Dataset uuu a aun ay uu yapa Sy uya yaaa Bees 24 How to use more than one covariate oooonccnccccnncnonnnoncnncnnnnnononanananonanan nono naar n cnn cnn ono esteen tetes nano nnnnanncaccacnnns 26 Getting Started Download the zipped file and unzip to a folder Make note of the folder s path e g C Users John Doe SAS Option as you will need this information later Do not change the names of any of the options A flow diagram on the next page explains the logic of the option Prior to running the option your data file must be formatted for META LY oF Sp ttr tr 00 cr 70 Tp T0 suondo suondo L I suondo nu mqns nu uiqns nu mqns 0 wed nun MEY ezAqeuy ZI EnSIA JO IZMENSIA H nu l i IJLLLAOD AJPLIBAOD AJPHBAOD IJPLIBAO Anaq mom gg a aaou moyga 99198 T YA 29787 ea I nun y amao e q T lgupAoO e Aq Z ajeneaoo e Aq T ayeneaod e Aq aerpaos suysnipy Mop Sunsnipy IM 3unsn py now 3unsn py WI pares I
30. s for the MRV among locations e Eigenvalues for cluster analysis e Eigenvalues for the PCA OUTPUT TO FILE e The variance components and broad sense heritability of the MRV by locations will be stored in Heritability Yield for Testing and Deleting Locations Option 11 csv e The phenotypic correlations among locations for the MRV will be stored in Phen Corr for Yield Among Locations Option 11 csv e The genetic correlations among locations for the MRV will be stored in Genetic Correlations for Yield Among Locations Option 11 csv e A dendrogram of the cluster analysis will be stored in Tree Yield Lattice Option 11 CGM 12 e The plot of the first 2 principal components will save to PCA Yield Lattice Option 11 CGM Option 12 BLUEs amp BLUPs Individual Analysis without Management Lattice This option will calculate BLUEs and BLUPs for the MRV BLUEs broad sense heritability LSD grand mean variance components and coefficients of variation for all of your variables it will correct the MRV for your covariate It will not separate out different management methods To run it type 1 and then 12 from the Data visualization vs analysis menu OUTPUT TO SCREEN e BLUEs and BLUPs for the MRV BLUEs broad sense heritability LSD grand mean variance components and CV for all variables in each location OUTPUT TO FILE e A summary file with BLUEs and BLUPs for the MRV BLUEs broad sense heritability
31. ss management conditions To run it type 2 and then 24 from the Data visualization vs analysis menu 16 OUTPUT TO SCREEN e Variance components and broad sense heritability for the MRV by location this will tell you if any of your locations is being suppressed due to poor heritability e BLUEs and BLUPs for the MRV and BLUEs for all other variables Nreps nlocs location variance genotype variance locationxgenotype variance residual variance grand mean LSD CV and broad sense heritability for each trait management combination OUTPUT TO FILE e A file with variance components and broad sense heritability estimates by location for the MRV will be stored in Heritability Yield for Testing and Deleting Locations Option 24 csv e A file with BLUEs and BLUPs for the MRV and BLUES for all other variables Nreps nlocs location variance genotype variance locationxgenotype variance residual variance grand mean LSD CV and broad sense heritability for each trait management combination will be stored in BLUEs amp BLUPs Combined Analysis By Management Option 24 csv Option 25 BLUEs amp BLUPs Combined Analysis Across All Locations RCBD This option will calculate BLUEs and BLUPs for the MRV BLUEs broad sense heritability LSD grand mean variance components and coefficients of variation for all of your variables it will correct the MRV for your covariate It will not break down the data by management condition but w
32. text describes the output of each data analysis option Any listed changes to the SAS code are optional they indicate commonly used but not required changes Option 1 Lattice Analysis with Covariate Adjustment Option 11 Genetic Correlations among Locations Lattice This option will calculate the genetic correlations between locations for the MRV adjusted by one covariate that you specify in the data input step It will also perform a cluster and PCA analysis based on the distance matrix To run it type 1 and then 11 from the Data visualization vs analysis menu CHANGES TO SAS FILE e Change the file Option 11 Genetic Correlations Among Locations Lattice sas e By default the dendrogram and PCA biplot are printed to a powerpoint editable CGM file If you wish for them to print to the screen instead o add a sign to lines 419 420 and 421 dendrogram and to lines 516 517 and 518 PCA biplot change the length of axis2 line 525 to 3 5 in from 7 0 in change the length of axis1 line 535 to 8 0 in from 9 0 in OUTPUT TO SCREEN e Variance components and broad sense heritability by location o Note if CodeForOut is 999 this means that the location s heritability was below your cutoff threshold and has been deleted from the analysis e ANOVA to verify the final locations retained in the data e Phenotypic correlations for the MRV with p values e Phenotypic correlations for the MRV without p values e Genetic correlation
33. ude FitStatistics ClassLevels run Be careful that you do not miss any places where Covar2 needs to be added Failure to modify the option properly will result in option failure or misleading results
34. variate so in the SelectCovariate menu selected option 1 25 Y SELECTCOVARIATE Command gt USING A COVARIATE Write 1 for With a Covariate Write 2 for Without a Covariate OPTION SELECTED 1 Next tell META where saved the SAS programs where want to store the results the name of data file and the names of my factors and variables SAS READING A EEE eon x TW File View Tools Solutions Window Help x Y y s m x Bo m o G z OG Command gt LATTICE WITH COVARIATE DATA ENTRY press ENTER to advance to the next variable CAUTION Write the correct name in the fields NOTE Any field can be corrected before pressing ENTER in Genotype Field Folder for Input Programs c path e path etc Folder Name for programs Folder to store results c path e path etc Folder Name to store results Input Data File Name Data File Input Main Response Variable MRU Input the Covariate Covariate mu Input Grouping Factor Name Grouping Factor Input Location Name Location Input Rep Name Replicate PR ss Input Block Name Block Input Genotype Name Genotype Pedigree ra Output Untitled El Log Untitled 2 Editor Untitled META Menus New Test Y READING CATest of META3 New Figures Next selected option 1 to do data analysis 26 LATTICE WITH COVARIATE OPTIONS Write 0 for BOX PLOTS amp HISTOGRAMS Wr

User`s Manual

Contents

Download Pdf Manuals

Related Search

Related Contents