Exact(6)
We used the 34 SNP genotype data and two inferred populations (K = 2) to assess the level of mixed ancestry detectable in the training set samples and to evaluate the degree to which outlying individuals shared patterns of allelic variability with the population sample as a whole.
By modelling the distribution of each of these summaries as a mixture, we inferred outlying individuals and excluded them from analysis.
Combining the data across participants reduced the impact of outlying individuals on the results and produced a model representative of the population.
Results both with and without outlying individuals removed are shown in Figure 3. Three non-coding SNPs, including our top two hits listed above, showed strong associations (P < 0.001) [see Additional file 2].
This method shows any outlying individuals with markedly lower assignment probabilities than the training set as a whole that can arise from above-average levels of mixed ancestry in the individuals.
This correlation was even stronger when removing the four outlying individuals (r = −0.68, P < 0.0001).
Similar(2)
The median-based approach to constructing our consensus classifiers is deliberately designed to be robust to the presence of outlying individual predictions.
Because quantitative analysis aggregates participants' judgements, outlying individual judgements are not represented.
Write better and faster with AI suggestions while staying true to your unique style.
Since I tried Ludwig back in 2017, I have been constantly using it in both editing and translation. Ever since, I suggest it to my translators at ProSciEditing.

Justyna Jupowicz-Kozak
CEO of Professional Science Editing for Scientists @ prosciediting.com