Figure 1.

Verification of the 35SNP test. (a) Structure algorithm run only on the complete set of data of wild‐living cats (35SNP_non‐ref_dataset), in the absence of any other reference data captive or domestic, reveals that results still correlate highly with scores generated for the full dataset; thus, any potential effect introduced by inbreeding within the captive population or biases in the reference data can be ruled out. (b) Independent assessment of 76 cats using 3,097 SNPs gleaned from the ddRAD pipeline reveals that the majority of estimates fall within the 90% CI obtained from the original test. (c) And the management decision taken on the basis of the 35SNP test would not appear to differ greatly where the estimate from the larger 3097SNP dataset taken to be the true hybrid score (see in text for details).(d) A graphical presentation of the absolute difference in values obtained by the 3097SNP and 35SNP test against the values obtained by the 35SNP test illustrates in more detail the expected variance in the results