PURPOSE: Machine learning techniques were used to identify which of 14 algorithms best predicts the genetic risk for development of proliferative vitreoretinopathy (PVR) in patients who are experiencing primary rhegmatogenous retinal detachment (RD). METHOD: Data from a total of 196 single nucleotide polymorphisms in 30 candidate genes were used. The genotypic profile of 138 patients with PVR following primary rhegmatogenous RD and 312 patients without PVR RD were analyzed. Machine learning techniques were used to develop statistical predictive models. Fourteen models were assessed. Their reproducibility was evaluated by an internal cross-validation method. RESULTS: The three best predictive models were the lineal kernel based on the Support Vector Machine (SMV), the radial kernel based on the SVM, and the Random Forest. Accuracy values were 78.4%, 70.3%, and 69.3%, respectively. The more accurate, although complex, algorithm uses 42 SNPs, whereas the simpler one uses only two SNPs, which makes them more suitable for routine diagnostic work. The radial kernel based on SVM uses 10 SNPs. The best individually predictor marker was rs2229094 in the tumor necrosis factor locus. CONCLUSION: Genetic variables may be useful to predict the likelihood of the development of PVR. The predictive capabilities of these models are as good as those observed with clinical approaches. These results need external validation to estimate the true predictive capability and select the most appropriate ones for clinical use.
PURPOSE: Machine learning techniques were used to identify which of 14 algorithms best predicts the genetic risk for development of proliferative vitreoretinopathy (PVR) in patients who are experiencing primary rhegmatogenous retinal detachment (RD). METHOD: Data from a total of 196 single nucleotide polymorphisms in 30 candidate genes were used. The genotypic profile of 138 patients with PVR following primary rhegmatogenous RD and 312 patients without PVR RD were analyzed. Machine learning techniques were used to develop statistical predictive models. Fourteen models were assessed. Their reproducibility was evaluated by an internal cross-validation method. RESULTS: The three best predictive models were the lineal kernel based on the Support Vector Machine (SMV), the radial kernel based on the SVM, and the Random Forest. Accuracy values were 78.4%, 70.3%, and 69.3%, respectively. The more accurate, although complex, algorithm uses 42 SNPs, whereas the simpler one uses only two SNPs, which makes them more suitable for routine diagnostic work. The radial kernel based on SVM uses 10 SNPs. The best individually predictor marker was rs2229094 in the tumor necrosis factor locus. CONCLUSION: Genetic variables may be useful to predict the likelihood of the development of PVR. The predictive capabilities of these models are as good as those observed with clinical approaches. These results need external validation to estimate the true predictive capability and select the most appropriate ones for clinical use.
Authors: Salvador Pastor-Idoate; Irene Rodríguez-Hernández; Jimena Rojas; Lucia Gonzalez-Buendia; Santiago Delgado-Tirado; Jose Carlos López; Rogelio González-Sarmiento; Jose C Pastor Journal: Clin Ophthalmol Date: 2017-05-22
Authors: Søren S Olesen; Carina Graversen; Stefan A W Bouwense; Harry van Goor; Oliver H G Wilder-Smith; Asbjørn M Drewes Journal: PLoS One Date: 2013-03-01 Impact factor: 3.240
Authors: J Carlos Pastor; Itziar Fernández; Rosa M Coco; María R Sanabria; Enrique Rodríguez de la Rúa; Rosa M Piñon; Vicente Martinez; Anna Sala-Puigdollers; José M Gallardo; Sara Velilla Journal: ISRN Ophthalmol Date: 2012-10-08
Authors: Rishika Chaudhary; Robert A H Scott; Graham Wallace; Martin Berry; Ann Logan; Richard J Blanch Journal: Transl Vis Sci Technol Date: 2020-02-21 Impact factor: 3.283