In this set of examples we will review the various query methods while looking at information about complex quantitative traits. In this case we'll be looking at QTLs in the RiceGenes database, so scroll down to RiceGenes and choose query.
Say you want to find all genomic RFLP loci in the RiceGenes database that are located on chromosome 4 that are related to a QTL. Which of the four query methods, Query by Example, Query Builder, WAIS search or fuzzy search, do you think would work best? You want locus records, you have a set of fairly specific restrictions, and all your criteria must be met (i.e. they would all be connected with a Boolean AND). A WAIS or fuzzy search would probably return you many more records than you actually want (like all the markers on chromosome 4, not just genomic RFLPs associated with QTLs). Either QBE or QB would be appropriate. Since your individual criteria are all AND'd together, QBE would probably be the simplest, so go ahead and select Query by Example.
Since we want Locus records returned to us, select the Locus class and hit the Query by Example button at the bottom. Now we need to translate our restrictions to the form. First we want genomic RFLP markers, so find Type and check the Genomic_RFLP box. To restrict this to chromosome 4 we will need to enter a pattern in the large box after the Location Map tag (since there is no separate field for "linkage group" or "chromosome number" or something like that). Remember in QBE the * is the wildcard, so you could enter *4* in the large box after Map. Finally, we want markers linked to a QTL (any QTL), so check the box in front of Linked_to_QTL. Remember, the small boxes are used to indicate "this field must exist". That should cover our search, so go ahead and submit this query.
This should return you a list of loci. (Notice they all begin with G or RG, which are indicative of genomic markers.) Select the one called RG788 and hit the button to display this record. We see this is indeed a Genomic_RFLP and it appears on chromosome 4 on three different maps. It has been linked to several different QTLs, from several different QTL studies. The naming convention in RiceGenes for QTLs is a lower-case q, followed by a one-character species identifier (r=rice, m=maize, etc), followed by a 2-5 character uppercase code for the specific trait measured (DLA=diseased leaf area, TN=tiller number, etc.), followed by a number indicating the chromosome and another number indicating the occurrence of a QTL for that trait on that chromosome.
Scrolling down a bit you see this marker has been mapped as both a framework and a low-LOD marker, and has also showed skewed segregation in one mapping study. Below that you see links to several alleles. Note that two of the alleles have a comment saying QTL allele effect. This means that those specific alleles were identified in some QTL study as having either a positive or negative effect on a trait. For breeding purposes, it is not sufficient just to know that RG788 was associated with a change in tiller number, for example. What you want to know is which allele at RG788 is associated with increased (or decreased) tiller number. Below the alleles are links to Map_data records which describe mapping studies that involved RG788, a link to the probe, and finally some remarks.
Scroll back near the top. To find out more about a specific QTL, follow the link to qr-TN-4-1 under Linked_to_QTL. This takes you to a QTL record, which gives you the approximate span of this QTL (from 31.4cM to 50.7cM on the map Rice-4). Each QTL study produces its own separate genetic map, which generally contains a subset of previously mapped markers. In RiceGenes, rather than include every individual QTL genetic map, QTLs have been overlayed onto the Rice consensus map. The Positive Contains markers listed in the window below were those markers from the QTL study which were shown to be associated with the trait Number of Tillers (but RG788 was the most significant locus).
Scrolling down you see this trait was evaluated in the Greenhouse. The significant locus, RG788, had a LOD score of 5.2. That means that the scenario in which RG788 is associated with number of tillers is 10 raised to the 5.2 power more likely than the scenario that it is not. In other words, the LOD score gives you an indication of how confident you can be in this association, the higher LOD conveying more confidence. The percentage of the variation in the number of tillers which could be explained by the allele present at RG788 is 10%, and the average difference in the number of tillers between the group of individuals homozygous for the parent1 allele at RG788 and the group homozygous for the parent2 allele is 1.7 tillers.
The Remarks give you interesting notes about other traits with which this marker is associated (such as root thickness) and also which allele was associated with an increase in tiller number - in this case it is the allele coming from the parent named CO39. Go back to the Locus record.
If you want more background on this QTL study, where it was conducted, what traits were measured and so on, follow the link to in_study QTL-Root-CO39/Moroberekan (this is a Map_Data object, telling you it is for a QTL map on root traits and the parents of the mapping population are CO39 and Moroberekan). In the Map_Data object you can find who to contact for more information, and an image which will show you the resulting genetic map and QTL locations as determined by the study. The remarks go into detail on the mapping population (this one was skewed toward the CO39 parent). The plant numbers are needed if anyone wants to map additional markers or traits using this mapping population, and they must line up with the plant scores as given later on for each locus and trait. The remarks also describe how the different traits were measured and how the QTL analysis was conducted.
Following the remarks are links to specific traits that were measured, the QTLs detected in the study and the various linkage groups. Following that are the loci used in the study along with the score (genotype) at that marker for each of the plants in the mapping population. After all the loci are listed, the phenotypic traits are listed, along with their plant scores.
Scroll to the portion of the record where the Trait field is located (right after the remarks). In RiceGenes, traits are organized in a hierarchy, with the highest (most generic) traits beginning with a "*". If you look at *Abiotic stress, you see all the specific traits that have been studied that are components of some abiotic stress. Also note that abiotic stress QTLs are assigned the color DARKGREEN, so when you see QTLs displayed graphically, all dark green ones will be abiotic stress QTLs. Go back to the Map_Data record and look at the trait for Total Root Number. Here you see the studies that have measured total root number, the higher level traits of which it is a component, and a list of total root number QTLs.
Go back to the Locus record for RG788. Scroll down to the Allele field and follow the link to RG788:DraI-A, which has been denoted as having a QTL effect. In the allele record we see that this allele is defined as a 7.3 Kb band when using the RG788/DraI probe/enzyme combination. Looking at the Trait_effect you see this marker has been implicated in a number of different QTL studies, and this particular allele (the 7.3 Kb allele) has been associated with decreased blast resistance, increased tiller number, decreased root dry weight per tiller, and so on. Following the trait_effect is a list of germplasm accessions which do not carry this allele (absent_in), followed by a list of those that do carry this allele (present_in). This is valuable information for a breeder, since you have an indication of what the effect will be if you select for this allele, and also a list of varieties that do/don't contain that allele. Many of those germplasm records will also have links into the GRIN database, where you can look at other traits as well.
Finally, to see how QTLs are shown graphically in RiceGenes, scroll to the top of the Locus record and hit the [view graphic]. RiceGenes makes extensive use of visual displays, many of which are not shown to their fullest using the WWW interface. If you find the visual displays useful, you may want to consider installing a version of RiceGenes on a local computer and using the ACEDB interface instead of the WWW interface. The graphic display you now see shows a set of colored bars on the left. These are QTL regions. Remember that the color code for Abiotic Stress QTLs is dark green, and you see several of these here on chromosome 4. Because we were looking at locus RG788 when we clicked the [view graphic] it has highlighted this locus (unfortunately, the locus is out of the display region on the right, so you may not be able to see it) and has also highlighted in pale blue any objects that are related to RG788. In this case, a number of different QTLs and their labels have been highlighted - all these QTLs identified RG788 as an significant marker. If you click on any of these highlighted QTLs, you will see the textual QTL record. The thin vertical lines to the right of the QTL names are homoeologous segments, which will be discussed in the next lesson.
That is the end of this example. You can continue to browse around in RiceGenes, return to the tutorial and try other examples, or go on to the next part of the lesson. Some WWW browsers may have opened this example in a NEW browser window (see if you have 2 instances of your browser running). If your browser has done this, close this instance of the browser to return to the tutorial, otherwise use the BACK or GO commands to return to the lesson.