[ Modélisation des Facteurs de Risque Génétiques dans l'Accident Vasculaire Cérébral Ischémique ]
Volume 6, Issue 1, August 2014, Pages 56–62
Khalid Balar1, Sellama Nadifi2, Khalil HAMZI3, and Bréhima DIAKITE4
1 Laboratory of Human Genetics and Molecular Pathology, University Hassan II, Faculty of Medicine, Casablanca, Morocco
2 Laboratory of Medical Genetics and Molecular Pathology, Faculty of Medicine and Pharmacy, Hassan II University, Casablanca, Morocco
3 Laboratory of Human Genetics and Molecular Pathology, University Hassan II/ Faculty of Medicine, Casablanca, Morocco
4 Laboratory of Human Genetics and Molecular Pathology, University Hassan II/ Faculty of Medicine, Casablanca, Morocco
Original language: French
Copyright © 2014 ISSR Journals. This is an open access article distributed under the Creative Commons Attribution License, which permits unrestricted use, distribution, and reproduction in any medium, provided the original work is properly cited.
In this article, we focus our work on the modeling of genetic risk factors on ischemic strokes occurred. To do this, logistic regression was widespread in our study. We proceeded in two stages: the first, we modeled the probability of the occurrence of ischemic stroke in an individual (i) based on genetic risk factors. Our sample consisted of 330 individuals aged at least 40 years, divided into 165 patients who had an ischemic stroke and 165 controls.
We applied the Wald test for all variables in the model one by one and we concluded to Reject H0, since the coefficients of our variables are not all zero.
In a second step, we studied the effects of these variables on the risk factors and then the effect of variables on Ischemic stroke to present the model equation.
We set a prediction threshold after specification test, that allows us to ensure the quality of the fit of the model and its degree of prediction, the proportion of people who have ischemic stroke is (50%). The results showed that 128 of 156 people with Ischemic stroke allowed a positive predictive value of 82%. We conclude that the prediction rate and the success rate of our model is 80.30 %, the results obtained with the «XLSTAT» software show a very good model with (sensitivity 78% and specificity 83%).
Author Keywords: Ischemic stroke, genetic risk factors, modeling, logistic regression and fit test.
Volume 6, Issue 1, August 2014, Pages 56–62
Khalid Balar1, Sellama Nadifi2, Khalil HAMZI3, and Bréhima DIAKITE4
1 Laboratory of Human Genetics and Molecular Pathology, University Hassan II, Faculty of Medicine, Casablanca, Morocco
2 Laboratory of Medical Genetics and Molecular Pathology, Faculty of Medicine and Pharmacy, Hassan II University, Casablanca, Morocco
3 Laboratory of Human Genetics and Molecular Pathology, University Hassan II/ Faculty of Medicine, Casablanca, Morocco
4 Laboratory of Human Genetics and Molecular Pathology, University Hassan II/ Faculty of Medicine, Casablanca, Morocco
Original language: French
Copyright © 2014 ISSR Journals. This is an open access article distributed under the Creative Commons Attribution License, which permits unrestricted use, distribution, and reproduction in any medium, provided the original work is properly cited.
Abstract
In this article, we focus our work on the modeling of genetic risk factors on ischemic strokes occurred. To do this, logistic regression was widespread in our study. We proceeded in two stages: the first, we modeled the probability of the occurrence of ischemic stroke in an individual (i) based on genetic risk factors. Our sample consisted of 330 individuals aged at least 40 years, divided into 165 patients who had an ischemic stroke and 165 controls.
We applied the Wald test for all variables in the model one by one and we concluded to Reject H0, since the coefficients of our variables are not all zero.
In a second step, we studied the effects of these variables on the risk factors and then the effect of variables on Ischemic stroke to present the model equation.
We set a prediction threshold after specification test, that allows us to ensure the quality of the fit of the model and its degree of prediction, the proportion of people who have ischemic stroke is (50%). The results showed that 128 of 156 people with Ischemic stroke allowed a positive predictive value of 82%. We conclude that the prediction rate and the success rate of our model is 80.30 %, the results obtained with the «XLSTAT» software show a very good model with (sensitivity 78% and specificity 83%).
Author Keywords: Ischemic stroke, genetic risk factors, modeling, logistic regression and fit test.
Abstract: (french)
Dans cet article, nous focalisons notre travail sur la modélisation des facteurs de risque génétiques sur la survenu des AVC ischémiques. Pour ce faire, La régression logistique a été largement répandue dans notre étude. Nous avons procédé en deux étapes, dans la première, nous avons modélisé la probabilité de la survenue d'un accident vasculaire cérébral ischémique chez un individu (i) en fonction des FR génétiques. Notre échantillon était composé de 330 individus âgés d'au moins 40 ans, répartis en 165 patients ayant fait un AVCI et 165 témoins.
Nous avons appliqué le test de Wald, pour toutes les variables du modèle une à une et nous avons conclu au Rejet de H0, puisque les coefficients de nos variables ne sont pas tous nuls.
Dans une seconde étape, nous avons étudié les effets de ces variables sur les FR et ensuite l'effet des variables sur l'AVCI afin de présenter l'équation du modèle.
On a fixé un seuil de prédiction après test de spécification qui nous permet de nous assurer de la qualité de l'ajustement du modèle et son degré de prédiction, la part des personnes qui ont AVCI (50%). Les résultats obtenus ont montré que 128 personnes sur 156 ayant un AVCI ont permis une valeur de prédiction positive de 82%. Nous concluons que le taux de prédiction ou le taux de succès de notre modèle est de 80.30%, les résultats obtenus avec le logiciel «XLSTAT» montrent un très bon modèle (sensibilité de 78% et spécificité de 83%).
Author Keywords: AVC ischémique, facteurs de risques génétiques, modélisation, régression logistique et Test d'ajustement.
How to Cite this Article
Khalid Balar, Sellama Nadifi, Khalil HAMZI, and Bréhima DIAKITE, “Modeling of Genetic Risk Factors in Ischemic Stroke,” International Journal of Innovation and Scientific Research, vol. 6, no. 1, pp. 56–62, August 2014.
