CIERA Analysis

Field Trial Analysis

CIERA can handle different types of field trial analysis. The analysis models, using R's lmerTest package, were set up from Dr. Fran Clarke's SAS models for analyzing lattice and RCB experimental designs.

Prerequisites

1) R (version 4.4.2) needs to be installed in order to run the Analysis module.

2) Install the following packages into R: lmerTest, emeans, emmeans, pbkrtest. This can be done through R Studio or by entering:
install.package("package_name") into the R console.

3) When launching the Analysis module, the system needs to set up an virtual Python Environment, which can take a minute or more. When you see a command prompt screen appear, please be patient!

Data Preparation - Input Files

1) Is is best practice to use data queried from CIERA, saved as a .csv file. However, data can come from any source but needs to be a .csv file. It is important that the following column headings are used:
* DESIGNATION - Line or germplasm name (character)
* ENTRYCODE - the entry code used for the germplasm (character)
* LOCATION - Location name or abbreviation (character)
* ZONE - Zone designation (character) - just set to '1' if only 1 zone
* REP - replication number (numeric)
* BLOCK - block number (numeric) (lattice designs only)
* PLOT - plot number (numeric)
* YLD KG/HA - Grain Yield in KG/Ha (numeric)
* MAT ACT - Maturity in actual days (numeric)
* HT ACT - Plant height in cm (numeric)
* LODG - Plant lodging in 1-9 scale (numeric)

2) Check each column carefully to make sure that there are no character text in the numeric columns, as this will cause the analysis to crash.

Random Complete Block (RCB) Design Analysis

Random Complete Block design has only REPS and PLOTS. Each germplasm is randomly assigned a plot within a rep.

Random Complete Block analysis

Inputs

1) File of plot data for Yield (YLD KG/HA), Height (HT ACT), Maturity (MAT ACT), and Lodging (LODG), if available at each location. File needs to include the REP, and PLOT at each location.
2) File of composite data for NIR-Protein Content (PROT NWG), Test Weight (TEWT), and 1000 kernel weight (MKWT).

Lattice Design Analysis

Lattice design has REP, BLOCK and PLOTS included in the experimental design. Each germplasm is randomly assigned a plot within a block and rep.

Lattice Design Analysis

Inputs

1) File of plot data for Yield (YLD KG/HA), Height (HT ACT) , Maturity (MAT ACT), and Lodging (LODG). File needs to include the REP, BLOCK and PLOT.
2) (optional) File of composite quality data for Protein (PROT), Test Weight (TEWT) and 1000 kernel weight (MKWT).

Saskatchewan Variety Performance Guide (SVPG) Analysis

Go to the SVPG site
All SVPG data is a Random Complete Block design , which has only REPS and PLOT. Each germplasm is randomly assigned a plot within a rep. The output will be slightly more customized to assist in development of the SVPG report, compared to the RCB analysis, which is more general.

SVPG Analysis

Inputs

1) File of plot data for Yield (YLD KG/HA), Height (HT ACT), Maturity (MAT ACT), and Lodging (LODG), if available at each location. File needs to include the REP, PLOT, and ZONE at each location.
2) (optional) File of composite quality data for Protein (PROT), Test Weight (TEWT) and 1000 kernel weight (MKWT).
3) Study metadata required: The entry code of the line (ENTRYCODE), the line name (DESIGNATION), REP, PLOT, name of location (LOCATION), and zone of the location (ZONE). Note that if the zone is undefined, just put a generic '1' in this column.
4) File of composite data for NIR-Protein Content (PROT NWG), Test Weight (TEWT), and 1000 kernel weight (MKWT).

Historical Yield Comparison Analysis

The analysis will take the current year's LS Means (plot data) or Means (composite quality data) and compare it against every location historically across years. The result is a comparison to see how the current year's yield compares to the historical yield.

Historical Comparison Analysis

Inputs

1) File of plot data for agronomic or composite quality trait for each location.
2) File of historical agronomic or composite quality trait for each location. File needs to include the YEAR for each location.

Run analysis

1) Enter the name of the trait column heading in your data (both the historical and current year files must have the same column heading).
2) Click on "Select a file with current year data" to find your current year's data file (.csv)
3) Click on "Select a file with historical data" to find your historical file of data (.csv)
4) Click on "Create an output file" and enter a filename and location to save the .xlsx analysis output.
5) Click on "Run Historical Comparison Analysis to start the analysis.
6) 2 files are created - one is your output file, which is set in step 4. The other is the output log from the python and R scripts that ran the analysis, for diagnostic purposes.