An investigation of the relationship between genetic and clinical data from amyotrophic lateral sklerosis (ALS) patients by using data mining techniques

Objective: Amyotrophic lateral sclerosis (ALS) is a motor neuron disease that controls voluntary muscle movement, affecting nerve cells in the spinal cord and spinal cord. Data mining is a discipline that allows extraction of unknown, meaningful sets of results, implicit or explicit in the database. This study was conducted with the aim of investigating whether there is a relationship between the clinical characteristics of individuals with ALS and mutation types, using data mining methods. Method: In this study, data from 65 patients diagnosed with ALS were used. The dataset contains 80 features, including clinical, genetic and demographic, for each patient.. R package program and software developed in Java programming language were used to examine the relationship between the features. On the data, hierarchical clustering and decision tree data mining methods were applied. Results: No relationship was established between environmental and demographic factors such as place of residence, gender, smoking status, and exercise status in ALS patients. Significant relationships were found in terms of clinical features. It has been determined that the decision tree method does not produce meaningful results on the data set. In terms of genetic factors, patients were grouped as SOD1 mutations and others. Significant results were not obtained in terms of ALS involvement and clinical features in mutation types. Again, statistically significant relationships did not appear in the analyzes made on demographic factors. Conclusion: Studies on data obtained in the field of health using data mining methods are quite common. The significance of the relationships between clinical findings in the study suggests that data mining methods can define various associations by creating more alternatives than the person can create. For this reason, it is inevitable that current or new methods of data mining become more important in the process, even though clinical evaluation is done for the diagnosis of the disease.