About the overgo plating process

The overgo plating process 
This page explains (briefly) the nature of the Overgo Plating experiments conducted at SGN. We also explain the terminology used on our Overgo statistics page in relation to the various stages of processing we have conducted.
About the overgo plating process 

Overgo plates have 96 wells, arranged in 8 rows by 12 columns. DNA sequence for one overgo probe (a sequenced marker from our maps collection) is assigned to each well. Wells are then pooled using the Soop program to produce 20 pools. (Equivalent to 8 row-pools + 12 column-pools.)

BAC plates are then run against these 20 pools for each plate to determine which BACs match a given pool. From this, the matches of BACs to probes can be inferred. A probe which successfully matches one or more BACs is said to "anchor" them to the Genetic Map and is thus referred to as an "anchor point".

To successfully anchor to a given probe, we require a BAC to:

  1. successfully match both its row-pool and column-pool, and
  2. not match any other pools on that plate.

In the case where BACs matched more than one row and one column on a given plate, we classify them as being ambiguous BACs. These are stored separately and are not shown on our physical map.

The FPC process 
Meanwhile, working from the other end, our collaborators at the Arizona Genomics Institute are using Fingerprint Contigging (FPC) techniques to assemble the BAC collection into contigs. The BAC <--> map position matches generated by our overgo plating experiment are used to inform the FPC process and improve the contigging conducted there.
In silico processing of BAC anchoring 

After establishing the unambiguous BAC <--> probe associations, a further stage of in silico processing was done to check for the plausibility of BAC matches. Our initial "plausible set" contains all unambiguously anchored BACs. We then remove BACs from this set which fail to meet certain criteria.

In general, we expect a given BAC to only match up to a portion of one chromosome. Thus, as a first step, we drop from our plausible set all BACs which are anchored to markers on two or more chromosomes. [N.B. - Once data from all plates are in, it is our intention to conduct more complex analysis. Specifically, if a BAC has multiple anchor points within a tight range on one chromosome and only a lone anchor point elsewhere then we may reasonably conclude that that one match is the aberrancy. However, at this stage the depth of data necessary to drive this analysis has not been accumulated.]

Secondly, we expect BACs to be anchored largely to one portion of a chromosome, rather than randomly anchored down its length. Thus we require a plausible BAC to have a "walk" of anchor points, all of which are within a reasonable distance of one another. Arbitrarily, we have chosen a distance of 5.0 cM as the maximum distance between any adjacent anchor points for a given BAC. BACs which are in violation of this principle are dropped from our plausible set.

Finally, we consider BAC contigs (as generated using FPC) to be plausible if:

  1. All of their member BACs which are anchored lie on the same chromosome, and
  2. There is a "walk" of BACs down the length of the contig where no two anchor points for the contig are more than 5.0 cM apart.
Summary of terms 

This section gives a summary of the terms used on the Overgo Plating Statistics page. To view the numbers for any of the following, please refer to that page.

Terms are listed here in alphabetical order.

Anchor Point
An overgo probe (qv) which has successfully been associated with one or more BACs is said to anchor them to the map and is thus referred to as an "anchor point".
BACs are Bacterial Artificial Chromosomes. The same library of BACs is used for both the FPC contigging and the overgo plating process. The total number of BAC clones on the library plates is 129024, but not all clones are valid and so the number of BACs reported is lower.
Empty wells
Overgo plates (qv) typically have 96 wells. In order to facilitate clear matching between one probe in a given well, the dispersal of markers on the plates has been deconvoluted so that no two markers sharing a pool lie within 5.0 cM of one another on the Genetic Map. A consequence of this is that it has sometimes been necessary to leave some wells empty on given plates in order to ensure a thorough dispersal. Thus, not every overgo plate contains 96 probes. Instead, some wells may remain empty. The total number of probes + empty wells for a given plate should add up to 96.
Hitting the plates
BACs which matched at least one overgo pool (corresponding to either a row or a column) on at least one of the overgo plates are said to have "hit the plates", in that they have at least been found to in some way match the sequences of the genetic markers known to SGN.
Overgo Plates
Each overgo plate contains up to 96 probes, arranged in 8 rows and 12 columns. To date, 16 plates have been processed.
Overgo Probes
Each probe refers to one sequenced marker found in the SGN tomato maps collection. Each marker is placed once on the overgo plates we designed and counts as a probe. If a probe successfully associates BACs to the map then it is said to be an anchor point.
Plausible BAC locations
The overgo plating process may produce multiple associations to probes for a given BAC. If those probes all lie on the same chromosome, within a relatively well clustered region, then we say the BAC is plausibly anchored to that chromosome by those anchor points. The criterion for "well clustered" is that no two consecutive anchor points are more than 5.0 cM apart. An FPC contig whose plausibly anchored BACs all lie on the same chromosome is said to be a plausible contig.
Unambiguous matching, a.k.a. plausible matching

Ambiguity refers to the match between a given BAC and a given overgo plate. If the BAC matches exactly one row-pool and one column-pool, thus specifying a clear match to one probe on the plate, then it is taken to be unambiguously anchored to that probe. Otherwise, the set of possible probes that it could be matched to is the set of ambiguous matches accorded to that BAC.