Skip to main content
. 2014 Dec 16;3:e04147. doi: 10.7554/eLife.04147

Figure 6. Integration of hedonic valence and odor concentration into ORDs.

(A) Response indices of wild type flies for all odors at median and high concentrations. Odors are sorted from highly aversive (−1, red) to highly attractive (+1, green). (B) 3D-scatter plot of median Ca2+ responses of all odors based on the three ORDs. Odor-dots are labeled due to their RI shown in (A). Same odors at different concentrations are connected with a line: the dot at the end depicts 10−2, the centered dot 10−4, and the end of the line 10−6. Attractive and aversive odor representations form separate clusters. (C and D) Left, schematic LH outlines with colored ORDs corresponding to data on the right. Correlation score r (upper right corner) between median activity and measured RI in T-maze experiments or odor concentration, respectively, with significance denoted below. Student's t test, *p < 0.05, ***p < 0.001. (E) Complete correlation matrices for Ca2+ response patterns of OSNs in the AL (left) and iPNs in the LH (right). The odors are arranged according to single linkage clustering of the LH activity patterns. Heatmap color-code refers to the correlation distance scale bar on the right. Correlation distance is defined as 1 − r, where r is the Pearson correlation coefficient between the response patterns of two odorants. Odor letters are color-coded according to hedonic valence; 10−6 RI values are labeled in grey (complete list right hand).

DOI: http://dx.doi.org/10.7554/eLife.04147.015

Figure 6.

Figure 6—figure supplement 1. Odor valences determined with three different behavioral assays.

Figure 6—figure supplement 1.

Odor-evoked behavioral responses of wild type flies for the 14 odors used in this study determined by T-maze assay, trap assay and the FlyWalk. The color denotes an attractive (green), aversive (red) or a neutral (light yellow) behavioral response. N/T, not tested. The majority of odors yielded similar results independent of the behavioral assay used. In a few cases an attractive odor evoked a neutral response (i.e., no response), but never induced an aversive response in another assay.
Figure 6—figure supplement 2. Calcium responses of OSNs.

Figure 6—figure supplement 2.

(A) Representative glomerular Ca2+-responses of OSNs in the AL for a subset of odorants at three concentrations. Scale bar to the right. Control (mineral oil) recordings are shown additionally as full false-color coded images. (B) Glomerular AL atlas used for glomerular identification. (C) Median Ca2+-activity traces of all glomeruli for all odorants at the three indicated concentrations. Scale bar and control measurement in the center. Odor application is indicated by the grey bar below the heatmaps (n = 6–7).
Figure 6—figure supplement 3. Correlation matrices for odor-evoked responses in the AL and LH.

Figure 6—figure supplement 3.

Complete correlation matrices for calcium activity patterns of OSNs in the AL (left) and iPNs in the LH (right). The odors are arranged according to single linkage clustering of the AL activity patterns. Heatmap color-code refers to the correlation distance scale bar below each matrix. Odor letters are color-coded according to hedonic valence; 10−6 RI values are labeled in grey (complete list right hand).