BA Milestone 9

View plot numbers, entry numbers, entry names, and germplasm codes in SOA job outputs and result files.

When viewing () job outputs of single occurrence analysis (SOA) requests, the Table of Residuals now show new columns for plot numbers, entry numbers, entry names, and germplasm codes. The Predictions tab also display the last three. These are also featured in downloaded () analysis result files. These improvements allow for easier identification and data analysis.

Downloading SOA analysis result files are improved.

In the ARM data browser, a SOA request’s status will not be set to CLOSED until the analysis result files are ready to be downloaded. Once you click on the download icon (), the zipped file should be available on your device.

Multi-occurrence analysis (MOA) requests are now supported.

In the ARM data browser, the Request Analysis drop-down menu now includes the option to request MOA. The following are featured:

1. Experiments tab. This lists all CLOSED SOA job inputs that you can select. Expand (>) each row to show all job IDs for each analyzed trait. The minimum requirement to proceed with the MOA request is to select at least two (2) job inputs under the same trait from different occurrences. Doing so enables the Next button. However, you can still add single job inputs from different traits to be included in the MOA request. This will be permitted by the system, but a notification will be provided once you click Next.

   

2. Options tab. This houses the options available for you. Specifically, you can:

   1. Download SOA result files only. This option lets you extract BLUEs and Analysis Metrics in a CSV format. Selecting either or both result files enables the Submit button, which you can click to confirm your download and redirect you to the ARM browser. As of now, BLUPs is not yet supported.

      

   2. Request for MOA [2nd step] only. This option lets you create one (1) MOA request. You can define the H2 threshold parameters before clicking Submit, which then redirects you to the ARM browser. Once ready, job results can be viewed () for each analyzed trait.

      

   3. Download SOA results and request for MOA [2nd step]. This lets you proceed with both options above. Simply Submit your request to be redirected to the ARM browser and wait for the analyses to finish.

      

3. Job outputs. On the ARM data browser, the action buttons are enabled once the job outputs for your MOA request is ready. You can view () the results page, where a list of jobs is shown at the leftmost section, while job outputs are at the right. In this version, the job outputs supported are the Analysis Metrics and Predictions. You can also download () these on your device.

   

Bug fixes

The following reported issues are resolved. You can now:

1. create an analysis request without encountering errors,

2. download SOA analysis reports (.xlsx) and MOA predictions_BLUEs results, and

3. correctly exclude occurrences with no quality-checked (QC) traits and whose trait data have a QC code of bad (B) in the SOA Experiments tab.

Viewing data in PDM is improved.

This version improves on the following features in the main data browser and the occurrence data browsers. Specifically, you can now:

1. View occurrence statuses. In addition to other basic information, the PDM data browser now also displays the occurrence status. This can be viewed after expanding (>) an experiment to see its occurrence/s. To recall, the PDM data browser only lists occurrences with the following statuses: PLANTED, TRAIT DATA COLLECTED, and TRAIT DATA QUALITY CHECKED.

   

2. View more experiments and occurrences. You can now view all experiment types in the PDM data browser. As such, trait data from all experiments and occurrences can now be inspected using PDM. Previously, only Breeding Trial (BT) experiments were available for trait data quality checking.

   

3. Switch to view another occurrence from the same experiment. You can now easily jump from viewing the trait data of one occurrence to another from the same experiment. Simply select another occurrence to display using the new drop-down menu just above the subtabs and wait for the new trait data to be loaded. This feature is also applicable when generating the box plot, histogram, scatter plot, and heatmap graphs. Take note that the drop-down menu is disabled for experiments with only one occurrence.

   

Management of transactions are enhanced.

The following features surrounding the management and saving of transactions are upgraded:

1. Save only BT experiments, except those using OFT stage. Saving (💾) and sending the transactions for analysis to ARM still only applies to BT experiments whose stage is not On-Farm Trial (OFT). This is related to the improvement on viewing all experiment types in the PDM data browser. A warning message is shown on the Save Transaction modal window to inform users of this limitation when trying to save the transaction of an unsupported experiment type and stage. Similarly, the Analyze button will remain disabled for these unsupported cases.

   

2. Save transactions with or without selecting traits. In the Save Transaction modal window, a new description is provided to inform you that you can now Save your transaction with or without selecting trait/s. Both saved changes are reflected in Core Breeding’s Experiment Manager. Previously, you can only save a transaction after selecting trait/s. This improvement affects the quality control (QC) codes of the saved data (💾), specifically for questionable (Q) data. This means that in each saved transaction:

   1. questionable (Q) data is converted to good (G) for selected traits, while

   2. questionable (Q) data, and all other data QC codes, are retained for unselected traits.

      

3. Revised criteria are now applied when marking traits as QCed. Trait data columns with at least one (1) good data and no questionable data can now be marked as quality-checked (QCed), provided that the user explicitly does so using the Save Transaction modal window. Additionally, traits whose data are all or have a mix of questionable, suppressed, bad, and missing cannot be marked as QCed.

   Previously, trait columns can be QCed if:

   1. all of its data are not missing (NA),

   2. it has no questionable data, and

   3. has been explicitly marked as QCed.

      

Uploading of sparse testing randomization files generated outside EBS is now supported.

In EBS, sparse testing is a way of randomly allocating test entries to different occurrences. Groups of occurrences with specific protocols are called parameter sets. These configurations include the trial designs applied to the parameter sets.

In this version, the backend features to support the upload of randomization files for sparse testing are now enabled in the Core Breeding domain’s Experiment Creation tool. This allows users to submit their own randomization files whose parameter sets use the Unspecified design and all other EBS-supported trial designs except IWIN and OFT, subject to validation rules.

Bug fix

You can expect to encounter no errors when:

1. validating a non-sparse testing file in Core Breeding’s Experiment Creation tool, and

2. generating the randomization for sparse testing, where the total number of plots created are now less than or equal to the specified NRep per entry.

RBAC is now supported.

The Breeding Analytics domain now implements role-based access control (RBAC), where users can only use tools that they have permission to access. These permissions are based on the roles assigned to their profiles in the Core System domain.

Bug fix

In the Table of Residuals, the bug causing the column filters to reset after selecting checkboxes is resolved.

Remain in PDM after all traits are QCed.

After saving () and marking all traits as quality-checked (QCed), you can now stay in PDM to further inspect your occurrence’s trait data. Once done, you can click the Analyze button to proceed to the ARM tool manually. Previously, users were automatically redirected to ARM once all traits have been QCed.

Bug fix

You can expect to encounter no errors when:

1. viewing accurate X, Y, and Z values in the heatmap graph, and

2. processing all QCed traits that were saved.