GEMDB»CONFLICT

Description

• Use the "Conflict Finder" to identify replicate samples with conflicting allele values, e.g. to identify "drift" between runs.

How to use this option

1. Go to SEARCH»GEMDB and select the data set to be analysed.
2. Mark the data set with a "Sample Tag" or with a "GEMDB Tag". If using a "Sample Tag", all the marker alleles for the selected samples are going to be analysed. If using a "GEMDB Tag", only alleles for markers that were included in the tagged GEMDB search are going to be analysed. (Please also refer to the section on tagging search results in GEMDB»SEARCH.)
3. Navigate to GEMDB»CONFLICT and select the tag grouping the samples (and markers, if applicable) from the drop-down box.
4. Choose from the conflict finder options (available from release 55):
   1. "Peaks on separate row" or "Peaks on same row"?
      • When multiple marker loci are distinguished with letter suffixes (e.g. hv0362a, hv0362b), conflict finder can either consider the alleles for the different loci separately ("Peaks on separate row") or it can ignore the letter suffixes and compare all allele values across all the marker loci ("Peaks on same row").
      • Only choose the latter option if there are concerns about inconsistently chosen letter suffixes.
      • See below for an example.
   2. "Conflicts per table cell" or "Alleles per table cell"?
      • "Conflicts per table cell" lists the alleles detected for a given sample in one cell of the results table. You find conflicting allele scores by comparing the string of entries between different cells (under column headings "Conflict 1", "Conflict 2", etc.).
      • "Alleles per table cell" shows the smallest alleles detected for all the samples under column heading "Allele 1", the next biggest alleles detected for all the samples are listed under column heading "Allele 2", etc. You find conflicting allele scores by comparing the entries inside a cell. If one or more cells on a row have two entries, followed by a cell with only one entry, this is an example where the conflict consists in one of the samples having more alleles recorded than the other sample.
      • See below for an example.
   3. Sort results by "Sample" or by "Marker" names?
5. Press the ANALYSE button.

The GEMDB»CONFLICT output

• GEMDB»CONFLICT will search the selected dataset for samples with identical names that were analysed with the same markers. For these, it will compare the converted allele values. If allele values for replicates differ, this is considered a conflict.
• By moving the mouse cursor over the conflict finder result table, information for every data point is displayed in a hover box, identifying the project, run name, etc. for a given allele entry.
• By clicking on an allele value, the GEMDB»BROWSE interface opens, filtered for the selected project » run » sample » marker » allele value. To return to the GEMDB»CONFLICT results with the browser's "back" button, you will need to "refresh" the screen (hit "F5" and click on the "retry" button).
• The conflict finder results can be downloaded. Note that when opening the downloaded delimited text file in a spreadsheet viewer, there are two tables: The first table displays the conflicting allele scores as shown in the GEMDB»CONFLICT output table (see also the example below). The second table contains all the supplementary information on a datapoint (user group, project, run name, etc...), as displayed in the "hover box" on the web interface.

GEMDB»CONFLICT input / output example

• Example input to illustrate how the different GEMDB»CONFLICT options influence the result.
• Hover over the allele scores in the example output tables for information on a data point. (Note that hover box entries are simplified for illustration purposes.)

Example input file

GEMDB»CONFLICT example output with options: "Peaks on separate row" / "Conflicts per table cell"

• "Peaks on separate row" option: hv5a and hv5b are considered separately. The input file has only one entry for hv5a, hence there is no "conflicting" allele score and hv5a is not listed in the output example.
• "Conflicts per table cell": Note that the alleles listed together in a table cell under column headers "Conflict 1", "Conflict 2", ..., are found on the same row in the "example input file".

GEMDB»CONFLICT example output with options: "Peaks on separate row" / "Alleles per table cell"

• "Peaks on separate row" option: hv5a and hv5b are considered separately. The input file has only one entry for hv5a, hence there is no "conflicting" allele score and hv5a is not listed in the output example.
• "Alleles per table cell" option: Note that the alleles listed together in a table cell under column headers "Allele 1", "Allele 2", ..., are found in the same column in the "example input file".

GEMDB»CONFLICT example output with options: "Peaks on same row" / "Conflicts per table cell"

• "Peaks on same row" option leads to hv5a and hv5b being considered as the same marker. As allele scores for hv5a and hv5b differ for Sample 2, hv5a is listed in the conflict finder output.
• "Conflicts per table cell": Note that the alleles listed together in a table cell under column headers "Conflict 1", "Conflict 2", ..., are found on the same row in the "example input file".

GEMDB»CONFLICT example output with options: "Peaks on same row" / "Alleles per table cell"

• "Peaks on same row" option leads to hv5a and hv5b being considered as the same marker. As allele scores for hv5a and hv5b differ for Sample 2, hv5a is listed in the conflict finder output.
• "Alleles per table cell" option: Note that the alleles listed together in a table cell under column headers "Allele 1", "Allele 2", ..., are found in the same column in the "example input file".

Back to contents

Links to other user manual pages

I'd rather try the tutorial...

• Get started with the GMT Tutorial
  Log in to 'GENica Marker Tools'