F2:3 population can be used as mapping population identify the QTLs for disease tolerance/resistance. However, DH lines/RILs would be the better choice for this study.
For mapping of genes F2 population is not enough since segregation pattern is the most concerned matter regarding to F2 population because F2 population is not completely homozygous that's why we always need nearly 100% homozygous population like RIL, DH etc need to more authenticate our data. Yes I agreed with M A Javed, You can take F2 derived population like F2:3.
As others have mentioned, using a DH or RIL population would be best for the QTL analysis. Since you have the F2s available, there are a few courses of action available to you. Firstly you could set up a single seed descent (SSD) from your F2s to produce a RIL population. Target a population of at least 300 individuals (Setup extras eg 350 to make sure you achieve your target population size). Descend the seed to at least F5 (or F7 if you want to be cautious) and then evaluate your RIL population and conduct the QTL analysis. You can speed up the SSD process by using constant lighting (ie 24 hrs/day) and small pots. Secondly, you could evaluate the remaining F2 seed for spot blotch resistance and use quantitative genetics formulae to estimate the number of effective resistance genes in the population based on their segregation ratios etc. If the population you have is suitable for use in a breeding program and the resistance appeared to be polygenic and additive you could select the resistant types from the F2 evaluation and continue selecting them through the generations (until about F5 when you could treat them as fixed lines). If the resistance appeared to be monogenic or perhaps two genes of large effect, selection through the generations may be unnecessary because you can identify those type of resistant lines when you characterise the RIL population. Good luck with your research.
U. Kumar with colleagues have identified 4 QTLs responsible for 62.9% of phenotypic variation in resistance to Bipolaris sorokiniana. One QTL that explained 22.7% of phenotypic variation was identified at 1 of 3 years of testing only. So, be prepare to study really large population and large number of seeds of F3-F5 population or >F5 lines for your purpose. Good luck to you in your hard work!
For disease resistance, a major gene involves in controlling the disease. if this the case in your population (see frequency distribution of disease) then you may conduct QTL analysis using F2 pop.
The most informative mapping population is F2 generation but in this population you can not replicate your data. Each individual in F2 population is independent.
You could extract the DNA from the F2 lines in the first SEASON and run your primers for the whole population (Minimum 100 lines). By this way you will finish the laboratory work. In the next season you could evaluate the family of each F2 line. You consider each plant in F2:3 heterogeneous inbred family as one replication.
If you are working on rice you could replicate F2 population for mapping as I have done in this paper ....
NO, because phenotyping has to be done across several seasons and environments in case of confirming disease resistance/susceptibility. This necessitates that mapping population has to be immortal type (i.e., amenable for maintenance and replication). Since F2 (segregating) population is not an immortal population, better to avoid using it to map disease-related traits/markers. Better, go for developing recombinant inbred lines (RILs) starting with F2 population and proceeding by single-seed-descent (SSD) approach.