TABLE 4.

Five examples of possible disagreements between PRS and validation data, when data harmonisation may be required. We illustrate five different situations in the table: Perfect agreement, labelling disagreement, strand flip, strand flip and labelling disagreement, palindromic (ambiguous) SNP.

		PRS summary data file		Validation data
		Effect allele	Non-effect allele	Effect allele	Non-effect allele
1	Expected scenario - perfect agreement	A	C	A	C
2	PRS and validation data disagree on labelling of effect allele	A	C	C	A
3	“Strand flip”	A	C	T	G
4	Strand flip and labelling disagreement	A	C	G	T
5	Palindromic	A	T	T	A