I have SNP data for association mapping in maize which is given in the base pair, I do need to convert it into cM distances to compare it with the previous QTL reports.
There is no fixed relationship between physcial and genetic distance, so you need to use empirical estimates from your species/population. If you have this data from markers you can generate a MareyMap to interpolate the relationship between genetic and physical maps.
http://pbil.univ-lyon1.fr/software/mareymap/
Article A graphical representation of genetic and physical maps: The Marey map
cM distances are based in recombination rates (1% recombination frequence is 1 cM). The transformations from bp to cM are quite difficult and only empirical, because changes between species (even populations), chromosomes or even chromosome location, as some sites have higher recombination rates in the same chromosome. If you really need it the only solution is the one provided by Maria Alcázar-Fabra
I agree with both of you that the relationship between physcial and genetic distance is merely empirical.
But, I would like to complile result of association mapping similar to article entitiled "The Genetic Architecture Of Maize Height" Table 2. in this they have given both the cM and bp distances.
If you read more carefully the paper you are referring (The genetic architecture of maize height...), you can only make the Table 2 data if you have both data on the specific part of the genome sequences of the organism and on the linkage analysis results between markers and the phenotype.
To my understanding, you need to have or develop the following from you QTL Mapping studies:
(a) the QTL locations for the target character,
(b) the markers flanking the QTL locations, and
(c) the distance between flanking markers to the QTL.
Moreover, you also need to have information about the genome sequences of the organism. If the genome sequences are available, then you can use the appropriate genome browser to locate the position of your flanking markers in the genome sequences. You can the measure the distance between the markers to the suspected genes existing between the flanking markers that may have associated functions to the characters. Once you are able to do so, then you can determine the bp distances between the markers and the associated genes.
By combining the cM data from linkage mapping studies and the bp distances from DNA sequence evaluation, then you will get something like you have mention in Table 2 of the maize height paper. The point I want to make is if you do not have both sets of data then it may not be possible for you to obtain what you desire.
I just hope this can add some points to the previous answers. Good luck with your research...