Peak number plot

Figure 8.4 Rotation function results P2 into crystalline ALBP. (a)Plot of the 101 best solutions to the rotation function, each peak numbered in the horizontal direction (abscissa). The correlation between the Patterson s of the probe molecule and the measured ALBP X-ray results are shown in the vertical direction (ordinate) and are given in arbitrary units. (b) Description of the rotation studies after Patterson correlation refinement. The peak numbers plotted in both panels (a)and (b) are the same. Reprinted with permission from Z. Xu et al. (1992) Biochemistry 31,3484-3492. Copyright 1992 American Chemical Society.

Peak frequencies of far infrared absorption spectra are plotted as dots along a linear scale. If the compounds have an odor in common or are specific attractants for a particular species of insect, the dots cluster at some places and avoid others. The number of dots in each 7-cm interval are counted and plotted to give the peak number plot. Lines drawn to standard deviations from the expected mean enable statistically significant favorable and adverse frequencies to be identified. 

Figure 2. Peak number plots for the Mediterranean fruit fly, or medfly, the Oriental fruit fly, and the melon fly. There is some overlap in the patterns, and many compounds with frequencies corresponding to the overlaps will attract two, or in some cases, all three species of insect.

Somewhat unexpectedly, the Peak Number Plots have provided a new insight into the mechanism of olfactory stimulation, for it is evident that the frequency-elements they reveal do not relate to any specific stimulus but rather to the frequencies around... 

Medfly Attractancy Pattern from the Peak Number Plot... 

ITsecond point suggested by the Table is the fact that where there are many favorable frequencies in the pattern, a randomly selected candidate is likely to have at least one that will approximately match one element of the pattern. This is no doubt the reason why 25.3% of the 2577 compounds tested by the U.S.D.A. attracted the medfly which has a seven element pattern. This contrasts with the fact that of 2618 compounds tested as attractants for the Mexican fruit fly, only 7.8% were effective (9). For this insect the Peak Number Plot shows only three favorable frequencies which makes it less probable that any given chemical will attract. 

Peak Number Plots for groups of compounds judged by expert human observers to have a fair degree of odorous similarity are shown in Figures 3, 4, and 5. Figure 5 includes the Plot for fifteen compounds whose odors are not related. These were compounds selected in 1966 by a Committee headed by Dr. L. M. Beidler as standard odor stimuli recommended for use in olfactory research (10). Their names and descriptions are shown in Table III. It is noticeable that in the absence of a common odor there is very little tendency to deviate from the expected mean frequency. 

Figure 3. Peak number plots for compounds having musky, bitter almond, and...

Figure 4. Peak number plots for the green, rose, and orange blossom odors...

Figure 5. Peak number plots for the musty and sweaty odors, and for 15 standard odorants which represent a variety of unrelated odors, thereby showing at most only one marginally significant deviation from the number of peaks to be expected in any randomly selected group of unrelated...

The even spacing is dramatically revealed by the "Peak Difference Plot" shown in Figure 6. This was developed by plotting differences between significant clusterings or gaps in 22 Peak Number Plots based on both human and insect evaluations, and counting the number of such differences in each 3 cm- interval. 

Note that differences between the positions of the peaks in the peak number plots are plotted as dots, and the number of dots appearing in each 3-cm interval are counted and plotted. The striking periodicity thereby revealed is interpreted as showing that the frequency sensitivities of the various biological receptors are evenly spaced. 

Taking 66 of the compounds used in constructing the Peak Number Plot for "musk" and rounding the observed peak positions to the nearest evenly-spaced value that was used in compiling Table II gives Table IV. 

The many qualifications made by expert "noses" were referred to above as constituting an obstacle to finding Peak Number Plots for particular odors. However, once a few such Plots have been developed it then becomes possible to take a fresh look at the comments and qualifications and begin to trace their origins in the respective molecular species. 

---