Here is the paper https://www.fsigenetics.com/article/S1872-4973(18)30248-5/fulltext#sec0170 "However, this mixture was designed to contain no more than four alleles at any locus to appear as a...
"However, this mixture was designed to contain no more than four alleles at any locus to appear as a two-person mixture if maximum allele count was used to infer the number of contributors [33]. In addition, only two (5A, 5B) of the four contributors were provided as reference samples. The profiles of the remaining two individuals in the mixture were not provided. Instead, a contrived profile of a reference not in the mixture (5C) was provided for comparison. The purpose of MIX13 Case 5 was to explore whether laboratories would consider this mixture too complex to interpret, and whether they would include the non-contributing reference profile (5C) and provide a matching statistic. The exact scenario of genotypes provided is highly improbable as has been noted"
"The Case 5 mixture was a very challenging test for interpretation. The mixture was made from an equal contribution of four unrelated individuals (i.e., in a 1:1:1:1 ratio) from the NIST population dataset. We selected the four individuals so that in the mixture no more than four alleles would be present at any locus (including the supplemental loci: Penta D, Penta E, D2S1338, and D19S433). On the surface, simply by allele count, the mixture would appear to be a two-person mixture. However, no laboratories in this study determined Case 5 to be a two-person mixture. Based upon the variability in peak height ratios and potential mixture ratios, labs typically reported that the mixture may have been from more than two individuals (see Supplemental File S2)."
"Because we designed the mixture to have a great deal of allele sharing among 17 loci (the 13 CODIS and the four supplemental loci from PP16 and Identifiler kits), we were unable to find a fifth unrelated person for comparison. Therefore, we constructed the 5C reference to share alleles among the four individuals in the mixture. We felt it was important to keep the same CODIS loci consistent for both PP16 and Identifiler kits since (a) this was done in MIX13 Case 1 through Case 4, and (b) we wanted to allow one-to-one comparisons at each locus between the kits. Interestingly, if we only made “four-person mixtures that looked like a two-person mixture” separately for each kit (i.e., only using 15 loci instead of 17 loci) we would find several “fifth” unrelated persons that could be compared, so it is unlikely that the contrived 5C reference versus a sample from a real person would have made a difference in the final analysis."
Here is the paper
https://www.fsigenetics.com/article/S1872-4973(18)30248-5/fulltext#sec0170
"However, this mixture was designed to contain no more than four alleles at any locus to appear as a two-person mixture if maximum allele count was used to infer the number of contributors [33]. In addition, only two (5A, 5B) of the four contributors were provided as reference samples. The profiles of the remaining two individuals in the mixture were not provided. Instead, a contrived profile of a reference not in the mixture (5C) was provided for comparison. The purpose of MIX13 Case 5 was to explore whether laboratories would consider this mixture too complex to interpret, and whether they would include the non-contributing reference profile (5C) and provide a matching statistic. The exact scenario of genotypes provided is highly improbable as has been noted"
"The Case 5 mixture was a very challenging test for interpretation. The mixture was made from an equal contribution of four unrelated individuals (i.e., in a 1:1:1:1 ratio) from the NIST population dataset. We selected the four individuals so that in the mixture no more than four alleles would be present at any locus (including the supplemental loci: Penta D, Penta E, D2S1338, and D19S433). On the surface, simply by allele count, the mixture would appear to be a two-person mixture. However, no laboratories in this study determined Case 5 to be a two-person mixture. Based upon the variability in peak height ratios and potential mixture ratios, labs typically reported that the mixture may have been from more than two individuals (see Supplemental File S2)."
"Because we designed the mixture to have a great deal of allele sharing among 17 loci (the 13 CODIS and the four supplemental loci from PP16 and Identifiler kits), we were unable to find a fifth unrelated person for comparison. Therefore, we constructed the 5C reference to share alleles among the four individuals in the mixture. We felt it was important to keep the same CODIS loci consistent for both PP16 and Identifiler kits since (a) this was done in MIX13 Case 1 through Case 4, and (b) we wanted to allow one-to-one comparisons at each locus between the kits. Interestingly, if we only made “four-person mixtures that looked like a two-person mixture” separately for each kit (i.e., only using 15 loci instead of 17 loci) we would find several “fifth” unrelated persons that could be compared, so it is unlikely that the contrived 5C reference versus a sample from a real person would have made a difference in the final analysis."