Análisis de Similitud en Representación de la Información en líneas celulares : Árbol de distancia correspondiente a la representación de líneas celulares basada en minería de texto.

Abstract

In the current research, the representation and comparison of cell lines play a crucial role in understanding the behavior and interactions of cell lines. This study implements an innovative approach to representing cell lines using text mining techniques, employing Principal Component Analysis (PCA) and Support Vector Data Description (SVDD). The main objectives were the extraction and processing of textual information from scientific literature, followed by its transformation into numerical representations, thereby developing a hierarchical clustering methodology. To achieve this, a dataset of cell lines was collected from sources such as Cellosaurus and PubMed. Python, along with libraries like pandas and scikit-learn, was employed for data processing, analysis, and modeling. The textual data underwent preprocessing, which included stemming and TF-IDF transformation, resulting in numerical feature vectors. PCA was utilized for dimensional reduction of the data while preserving its variance. SVDD identified atypical values, and the distances between cell lines were visualized through the construction of a dendrogram. The results revealed an innovative way to represent cell lines, allowing for the identification of groups and similarities among different lines. The dendrogram visually represented hierarchical relationships, providing insights into cell line clusters.