Table 1. Complete list of primary scores, Wald χ², degrees of freedom and P values from logistic models for single-plex and 12-plex SNP-IT reaction.

Score name	Score description	Single-plex model χ2, df, P value	Multiplex model χ2, df, P value
TIN	The free energy of the most stable structure of extension primer with itself	2.4, df = 1, P < 0.0001
Most stable 3′-end dimer	The free energy of the most stable dimer formed by annealing SNP-IT primer with itself. Only the dimers formed by last nine 3′-end bases are considered for this score. In addition a binary becoming 1 when no dimers are formed at the 3′-end is used for single-plex model	65.2, df = 1, P < 0.0001
TDN	The free energy of the most stable structure between extension primer and PCR product	28.1, df = 1, P < 0.0001	12.3, df = 1, P = 0.0005
Cumulative TDN	Free energies sum of all possible dimers (no loops) between extension primer and PCR product	36.6, df = 1, P < 0.0001	7.1, df = 1, P = 0.0077
Number of consecutive Gs in SNP-IT primer	Binary score becomes 1 when number of consecutive Gs in SNP-IT primer is six and greater	7.1, df = 1, P = 0.0078
Stability of last 9 bases of extension primer	The free energy of the SNP-IT primer 3′-end. In addition to this score single-plex model uses its cube approximation and multiplex model uses its square approximation:
	Linear term of stability of last 9 bases of extension primer	60.1, df = 1, P < 0.0001	284.1, df = 1, P < 0.0001
	Square term of stability of last 9 bases of extension primer		172.0, df = 1, P < 0.0001
	Cubic term of stability of last 9 bases of extension primer	32.1, df = 1, P < 0.0001
Number of C3 linkers	Number of C3 linkers in the extension primer		149.3, df = 1, P < 0.0001
	Binary score becomes one when number of C3 linkers is equal to 1	16.4, df = 1, P < 0.0001
	Binary score becomes one when number of C3 linkers is greater than 3	8.9, df = 1, P = 0.0029
%GC of PCR primers		29.7, d.f.=1, P < 0.0001	43.4, df = 1, P < 0.0001
Amplicon melting temperature			106.0, df = 1, P < 0.0001
	Binary score becomes one when amplicon Tm is <73°C		106.0, df = 1, P < 0.0001
Number of ambiguous bases	Number of ambiguous bases in amplicon	27.6, df = 3, P < 0.0001
Number of repeats	Number of repeats in amplicon		109.1, df = 1, P < 0.0001
	Binary score becomes one when number of repeats in amplicon is 8 or 9	22.0, 1, P < 0.0001
Amplicon structure around SNP site	The free energy of the most stable amplicon structure (with loops) containing 5 bases of extension primer annealing site and 2 bases upstream of primer annealing site		231.5, 1, P < 0.0001
Amplicon structure around PCR primer annealing sites	The free energy of the most stable amplicon structure (with loops) containing 5 bases of PCR primer annealing site and 2 bases upstream of primer annealing site. Only one the most stable structure between two PCR primers was considered		80.2, df = 1, P < 0.0001
Extension mix change date and 13 interactions with combinations of extension mix and last 3′-end of SNP-IT	This binary score was introduced for the single-plex model to reflect historical assay modification. New extension nucleotides were introduced at Orchid Bioscience into SNP-IT assay	491.6, df = 14, P < 0.0001a
Extension mix + last 2 bases at 3′-end of SNP-IT primer	24 different combinations were used for the single-plex model and 67 combinations for the multiplex model out of 96 possible. Only combinations which show significant correlations with the failure/success rate were considered. In addition, 3 interactions scores with extension mix change date score are used for the single-plex model	187.7, df = 24, P < 0.0001a	543.0, df = 67, P < 0.0001a
Extension mix + last 3 bases at 3′-end of SNP-IT primer	42 different combinations were used for the single-plex model and 56 combinations for the multiplex model out of 384 possible. Only combinations which show significant correlations with the failure/success rate were considered. In addition, 7 interactions scores with extension mix change date score are used for the single-plex model	208.2, df = 42, P < 0.0001a	255.5, df = 56, P < 0.0001a
Extension mix + last 4 bases at 3′-end of SNP-IT primer	34 different combinations were used for the single-plex model and 60 combinations for the multiplex model out of 1536 possible. Only combinations which show significant correlations with the failure/success rate were considered	101.8, p < 0.0001a	155.1, p < 0.0001a
Extension mix + last 5 bases at 3′-end of SNP-IT primer	3 different combinations were used for the single-plex model and 18 combinations for the multiplex model out of 6144 possible. Only combinations which show significant correlations with the failure/success rate were considered	9.3, P = 0.03a	53.6, P < 0.0001a
Last 2 bases at the 3′-end of the PCR primer + 1 amplicon base next to PCR primer annealing site	2 different combinations were used for the single-plex model and 7 combinations for the multiplex model out of 64 possible. Only combinations which show significant correlations with the failure/success rate were considered	5.9, P = 0.05a	36.0, P < 0.0001a
Last 3 bases at the 3′-end of the PCR primer + 1 base next to PCR primer annealing site	22 different combinations were used for the single-plex model and 24 combinations for the multiplex model out of 256 possible. Only combinations which show significant correlations with the failure/success rate were considered	91.1, P < 0.0001a	70.5, P < 0.0001a
Last 2 bases at the 3′-end of the PCR primer + 2 amplicon bases next to PCR primer annealing site	19 different combinations were used for the single-plex model and 18 combinations for the multiplex model out of 256 possible. Only combinations which show significant correlations with the failure/success rate were considered	67.4, P < 0.0001a	48.5, P = 0.0001a
Last 3 bases at the 3′-end of the PCR primer + 2 amplicon bases next to PCR primer annealing site	40 different combinations were used for the single-plex model and 49 combinations for the multiplex model out of 1024 possible. Only combinations which show significant correlations with the failure/success rate were considered	135.4, P < 0.0001a	128.8, P < 0.0001a
External TIN	The free energy of the most stable structure formed between marker extension primer and other extension primers in the multiplex cluster		15.0, df = 1, P = 0.0001
External TDN	The free energy of the most stable structure between marker extension primer and other PCR products in the multiplex cluster		39.5, P < 0.0001
External PCR TDN	The free energy of the most stable structure formed between any of two marker PCR primers and other PCR products in the multiplex cluster		55.4, P < 0.0001
Dispersion of amplicon melting temperature	The score is calculated as the sum of absolute differences between Tm of marker amplicon and Tm of all other amplicons in the multiplex cluster		16.3, P < 0.0001
Dispersion of amplicon complexity	The score is calculated as the sum of absolute differences between marker amplicon complexity and complexity of all other amplicons in the multiplex cluster		22.6, P < 0.0001

An empty cell for the score indicates that it was not included into the corresponding model.

^aThe difference in Wald χ² between the final model and the model without the set of binary variables, degrees of freedom and P value.