In the preceding report we completed the protocol for a thorough analysis
involving p, recombination values and** **a and b, the allometric coefficients
of the effects of genes *A* and *B* and the variance of all parameters.
To evaluate the potential value of the procedure, analysis was illustrated
in actual numbers with data from D.L. Langham (Genetics 25:88) involving
F2 data between *pd* and *tr* and standard markers in crosses
of maize with Durango teosinte. Searching for another batch of data with
the widest possible interest we analyse here backcross data involving translocations
*T wx* reported by E.G. Anderson, et al. (MNL 39:106-109, 1965). We
picked only those which are in the waxy and nonwaxy reciprocal translocations
listed in MNL 61:113, which are a standard set. The results are presented
in Table 1. The p values were calculated by the Product Ratio method, by
maximum likelihood and by our allometric method. Our method gives p values
between the others, nearer to the maximum likelihood, and the p values
are always smaller in the normal cases. But most important are the even
smaller variances, so that a much more exact pooling of different sets
of data can be arrived at by weighting the invariances.

There is no sign of allometrical effects indicated by the t values. In other words, semi-sterility affects equally all four classes a, b, c, d. The Product Ratio method exaggerates the strength of linkages, underestimating the error variances at the tight p values.

The mean of the 12 p values gives p = 5.91. In MNL 61:34-35 using our
whole set of genetricks we obtained a mean p = 7.54 for 12 estimates of
the standard *T wx* set, although between them only 2 are common.
Of these, in 10 cases we believe that, more probably than not, the correct
orientation was arrived at using information from some of our other reports.
Perhaps a more realistic comparison is made by taking out 9L.10;5L.14,
giving then p = 6.31 (it was deleted from the standard set). For the present
set S.17;L.69 should then be taken out, and now gives p = 3.29 (it should
also be discarded from the standard set). Taking into account also the
common segregation for *Krn* factors, we can in many cases increase
considerably the efficiency of the process of mapping as we have done in
prior works. We think the correct mapping of *T wx *(and *Krn*
factors) of utmost interest for greater marginal returns by getting the
second approximation to the correct position in one strike.

Table
1. In this table are analyzed the data with those *wx* T9 translocations
reported by Anderson et al. in MNL 39:106-109, 1965 that are listed in
the standard translocations from the Cooperation Stock Center in MNL 61:113.

Luiz Eugenio Coelho de Miranda and Luiz Torres de Miranda

Return to the MNL 62 On-Line Index

Return to the Maize Newsletter Index

Return to the Maize Genome Database Page