Table 2. The molecular nature of sequence variants and their affect on phenotypic variation.

QTL Pct Var	Intergenic	Downstream	Exon	Intron	Upstream	Coding (detrimental)	SNP	Structural variant	Indel
All	1.18**	0.71	0.7	0.79	0.67	0.79	1.00	0.84	1.04
<4%	1.21**	0.67	0.67	0.75*	0.63	0.74	0.99	0.69**	1.07
>4%	0.57**	1.05	1.28	1.43*	0.97	1.00	1.02	0.85	0.95
>10%	0.65**	1.32	1.59*	1.69**	1.32	2.13*	0.88**	1.69*	1.48**

The class of sequence variants and their position relative to genes influence the likelihood that they are functional, as predicted by a statistical method ⁴⁴. The table shows the ratio of variants that score a maximum negative merge log(P-value) to those that do not within five different genomic regions: intergenic, exonic, intronic and either 2 kb upstream or downstream of the gene. Ratios are also shown for four molecular types: SNPs, structural variants, insertion/deletions (indels) polymorphisms and SNPs predicted to be detrimental to the coding sequence of a gene. The QTL data used for this analysis were derived from the heterogeneous stock mice²⁴ generated from a cross between eight of the sequenced strains.

^*P < 0.05

^**P < 0.01.