Identification of Chiari Type I Malformation subtypes using whole genome expression profiles and cranial base morphometrics

Table 2 Sparse k-means clustering results ^a

Description	Individual clustering
	Dura	Blood	Cranial
Optimal k classes ^b	2	2	2
Class 1 (N)	24	25	21
Class 2 (N)	20	19	19
Optimal tuning parameter	68.27	81.71	3.66
N non-zero weighted features (%) ^c
Dura	11804 (100%)	NA	NA
Blood	NA	11804 (100%)	NA
Cranial	NA	NA	18 (75%)
Maximum weighted feature (weight)
Dura	MECOM (0.21)	NA	NA
Blood	NA	RPS7 (0.17)	NA
Cranial	NA	NA	BASTOREF (0.72)
Gap statistic ^d	1.152 ± 0.013	1.834 ± 0.014	0.487 ± 0.136
95% Confidence interval	1.126-1.178	1.807-1.860	0.222-0.753
P-value	<1.0E-10	<1.0E-10	1.62E-04

Abbreviations: N: number, MECOM: MDS1 and EVI1 complex locus, RPS7: ribosomal protein S7, BASTOREF: basion to reference line, NA: not applicable.
^aClustering results presented using individual datasets (Dura: dura gene expression data, Blood: blood gene expression data, Cranial: PF trait data).
^b44 individuals were included for the dura and blood gene expression individual clustering analyses; 40 individuals were included for all other analyses.
^c11804 gene expression probes and/or 24 posterior fossa traits (features) were used as input.
^dThe gap statistic ± standard error is presented. Gap statistics with an approximate p-value less than 0.05 are shown in bold.

ISSN: 1755-8794