Abstract Details

Title In Silico Predictions of KCNQ Channel Variant Pathogenicity in Epilepsy

Topic Child Neurology and Developmental Neurology

Presentation(s) P5 - Poster Session 5 (5:30 PM-6:30 PM)

Poster/Presentation
Number 7-049

Objective To identify a prediction model that determines if KCNQ channel variants in infantile epilepsy are pathogenic and to predict the severity of epilepsy based on location of the variants.

Background Mutations in KCNQ channels have been associated with early onset neonatal encephalopathy and benign familial neonatal epilepsy. With the increased availability of genetic testing, many patients are being found to have variants of unknown significance in several KCNQ channels. Currently there are no tools for the clinician to be able to determine if these variants are pathogenic or the severity of the epilepsy syndrome. There are several computer algorithms available to predict pathogenicity of variants, however their results do not always correlate.

Design/Methods Databases of reported KCNQ2 and KCNQ3 variants in patients with neonatal epilepsy were accessed. The variants were plugged into eight available prediction programs. Sensitivity, specificity, percent correct decision, and the classification accuracy were determined and compared for each program. Pathogenic mutations were then mapped onto the topographic representation of the channel and mathematical models used to predict severity of the phenotype.

Results In patients with epilepsy, PROVEAN prediction algorithm was accurate 92% (sensitivity 0.90 and specificity 0.94) of the time and the next best algorithm was correct 86% of the time. PROVEAN met the American College of Medical Genetics standard for determining pathogenicity in KCNQ channels. The KCNQ channel variants showed no distribution that predicted severity of epilepsy but did show that regulatory domains of the channel are more likely to have pathogenic variants.

Conclusions These results show that the clinician can use a free online tool to predict whether a variant of unknown significance in KCNQ channels is likely to be pathogenic but more work is needed to be able to predict the severity of the disease.

Authors/Disclosures
David Ritter, MD, PhD (Cincinnati Children's Hospital Medical Ctr) PRESENTER	No disclosure on file
	No disclosure on file
Katherine D. Holland, MD, PhD (Cincinnati Children's Hospital)	No disclosure on file

��ɫ��

��ɫ��