Methods: REP-PCR Identifying the Source Groups of Water Isolates Using the Database (6-01-04)

Step 1. Create a library of known source isolates

From the main menu select identification/create new library

Enter a library name - for example 'sourceID'

Library name will appear on the screen under the 'libraries' heading

Double click on the library name 'sourceID'

The 'sourceID' menu will appear, under the 'fingerprint types' section, highlight Box and then select experiment/use for identification - a green arrow should appear next to the word Box

From the 'sourceID' menu select file/add new library unit

Enter unit name - for example 'chickens'

Library unit name will now appear in the 'units' section of this menu

Double click on the word 'chickens' to select this unit

The window 'unit-chickens' will appear, select all the fingerprints from chicken isolates in the main database window (selected entries will be marked with a blue arrow) then select edit/paste in the 'unit-chickens' window

The selected chickens will now appear in this window

Repeat procedure to create units containing isolates obtained from each animal source represented in the main database (cows, chickens, humans, dogs, etc.), close all windows and return to the main database

Step 2. Identifying unknown source groups using the library

Highlight the fingerprint entries in the main database obtained from contaminated watersheds and therefore from unknown sources (for example "Mississippi River" isolates), isolates from both known and unknown source are included in a single database

Select comparison/create new comparison, a comparison window will appear In this widow select 'Pearson's product-moment coefficient' similarity coefficients, we use this curve-based method to compare entries, click the position settings button to bring up the comparison settings menu, set all options to 1% (optimization, position tolerance, change toward end of fingerprint, minimum height and surface), these options will be used when comparing the fingerprints from the water isolates to the isolates from known sources, close all windows and return to the main screen (these options may be changed if you wish to calculate your comparisons using other methods)

Select Identification/Identify

Selected entries, answer yes to 'calculate quality factors'

An identification window will appear containing the selected river isolates and units with which they had the highest average similarity

Select show/maximum similarities- we are currently using maximum similarity values rather than average similarity values for source group identification

Next to each unknown the unit name containing the isolate with the greatest similarity is listed, along with the percentage of similarity between the two fingerprints

For example, Mississippi river isolate 122 most closely matches a goose fingerprint with 76% similarity, while isolate 123 most closely matches a dog fingerprint with 83% similarity, currently a maximum similarity value greaater than 80% is required for a positive identification, therefore the first isolate would remain unidentified and the second isolate would be identified as a dog, the 80% value allows identification of almost all of our unknown isolates