Behavioral signatures related to genetic disorders in autism

Table 6 Application of the SVM algorithm derived from the genetic disorder sample to the different AGRE datasets

Genetic disorder	AGRE0				AGRE1				AGRE2
	n	% assigned	Mean probability	SD probability	n	% assigned	Mean probability	SD probability	n	% assigned	Mean probability	SD probability
22q11DS	26	6.9	0.44	0.154	23	5.2	0.44	0.143	28	6.3	0.48	0.189
Down’s	1	0.3	0.25	NA	1	0.2	0.28	NA	1	0.2	0.27	NA
PWS	1	0.3	0.34	NA	5	1.1	0.30	0.131	5	1.1	0.33	0.072
SMC15	24	6.4	0.40	0.086	28	6.3	0.40	0.102	32	7.2	0.41	0.093
TSC	255	68	0.61	0.139	302	68.2	0.62	0.134	283	63.9	0.60	0.140
XXY	68	18.1	0.41	0.092	84	19	0.41	0.071	94	21.2	0.42	0.095
Total	375	100			443	100			443	100

% assigned, percentage assigned to the respective genetic disorder class; 22q11DS, 22q11.2 deletion syndrome; AGRE, Autism Genetics Resource Exchange; Down’s, Down’s syndrome; mean probability, average predicted probability; n, number of assigned cases to the respective genetic disorder class; NA, not applicable; PWS, Prader-Willi syndrome; SD probability, standard deviation of predicted probability; SMC15, supernumerary marker chromosome 15; SVM, support vector machine; TSC, tuberous sclerosis complex; XXY, Klinefelter syndrome.

ISSN: 2040-2392