Table 1 Overview and summary statistics for the five types of SNV-sets analysed and the number of significant loci identified with each of the seven SKAT models used.

Cis/Trans	SNV-set	No. of SNV-sets^a	No. of SNVs^b	Rare^c	No. significant loci for the different models^d
					1	2	3	4	5	6	7
Cis	Cis-Reg	405	61 [31–116]	47.0%	158	154^g	83	161	175	178	40
Cis	Cis-Flank^e	405	190 [118–307]	48.4%	148	148^g	80	151	167	176	45
Cis	Cis-CDS	405	8 [5–13]	49.1%	138	132^h	87	150	158	167	43
Trans	Trans-CDS	18,467	8 [5–14]	54.2%	19	26^h	6	21	24	26	3
Trans	Trans-Flank^f	18,467	229 [161–330]	51.3%	19	20^g,h	8	21	26	26	5

^aFor Cis-SNV-sets, each of the 405 autosomal SNV-sets were analysed only in relation to the encoded protein, whereas in the Trans-SNV-sets, the SNV-set (one for each of the 18,467 genes across the genome) was analysed in relation to all 414 proteins. The significance threshold was 0.05/3 Cis-sets/405 proteins/seven models = 5.88 × 10⁻⁶ for Cis, and 0.05/2 Trans-sets/414 proteins/18,467 SNV-sets/seven models = 4.67 × 10⁻¹⁰ for Trans.
^bMedian [interquartile range] of the number of SNVs in the SNV-sets.
^cFraction of SNVs and indels in the SNV-sets that were considered rare (MAF < 0.0239).
^dThe seven models are: model (1) Unweighted, model (2) CADD or Eigen weighted, model (3) MAF weighted—β(1, 25), model (4) MAF weighted—β(1, 5), model (5) MAF weighted—β(0.5, 0.5), model (6) CommonRare, model (7) Rare only, model. See method section for more information on the models and Supplementary Fig. 7 for information on the β-distributions.
^eGene ± 100 kb-regions up/downstream of each gene, filtered by Eigen >10 when analysed in Cis.
^fGene ± 100 kb-regions up/downstream of each gene, filtered by CADD or Eigen >10 when analysed in Trans.
^gWeighted by Eigen values.
^hWeighted by CADD values.

Quick links

Search