The statistic showed in the “Possible inconsistency window” concerns how many data points out of total data points (XX/YY) will be used in gap filling (read-across or trend analysis).
This statistic does not show how many chemicals will be used in gap filling stage; it is related to data points ONLY.
By default gap filling approach works with average values (in trend analysis and read-across). If one chemical has more than one data points they are recalculated and used as average value in the building the equation (in case of trend analysis) or calculation of average weight (in case of read-across). For instance in gap filling I should have 151 out of 162 data points as shown in the “Possible inconsistency” window, see Figure 1):
Figure 1 Possible inconsistency window
How to see these 151 data points, please follow the steps:
Once you are in gap
filling, go to Calculation options (1), then select All data points(2)
and click OK button (3) as shown in the Figure 2:
Figure 2. Calculation options
Once this option is selected, you will see data points distributed for a given chemical see Figure 3.
Fig. 3 Distribution of all data points across chemicals
Keep in mind that read-across is based on 5 nearest analogues with respect to log Kow. So not all 151 data points are taken into read-across prediction. In order to use all data points, you should follow the steps:
1. Open Prediction approach options tab and increase the number of neighbors as shown on Figure 4.
Figure 4 Prediction approach options