>>Colleagues: Allow me to give a slightly different perspective to this
>>problem and in the process probably repeat what others have stated. A
>>problem with the use of "genetic coefficients"  may be with the term
>>itself, which has an absolute connotation limited to the discipline of
>>crop simulation modeling. If one goes to a plant breeder and asks for the
>>genetic coefficients for a particular cultivar, a probable response will
>>begin with confusion over this term followed by a lengthy explanation by
>>the one asking the question, and ending up without the desired answer. In
>>addition, there may be the perception by some that if it is related to
>>plant genetics, it can't be easily changed, thus these coefficients are
>>fixed. However, these coefficients are determined by an empirical process
>>where development stages and key growth parameters are noted and the
>>model is run in an iterative fashion until optimal values for the
>>coefficients are reached which agree with the observed values. Perhaps a
>>better term for these coefficients might be "genotype trait coefficients".
>>         While this has been a very interesting discussion, what do we do
>> next to further the goal of improving crop simulation modeling by
>> incorporating genomics? It won't be easy. Here are two suggestions; the
>> idea is to keep improving upon these suggestions until we reach some
>> sort of a consensus for research directions. (There  may be other
>> approaches to address this issue that might work well. Make a comment to
>> the list and see what happens.) The first suggestion is from a modeling
>> perspective. If there is a robust process that is consistently well
>> simulated, perhaps one can look at the genetic mechanisms that control
>> this process to determine what is known about the genes controlling this
>> process. And if appropriate, how to quantify these relationships. The
>> other suggestion is to find a trait where the genes for this trait are
>> well documented and determine if appropriate algorithms can be
>> developed. Any of these approaches will require collaboration with those
>> who have expertise in genomics and probably other areas.
>>         Albert Weiss