Investigation of the Roles of CRIP1 and IFITM1 as a Transcriptional Marker to Identify Periodontitis with Neural Network
Jeewoo Lee *
Computer Science Division, STEM Science Center- 111, Charlotte Place/Englewood Cliffs, NJ -07632, United States.
*Author to whom correspondence should be addressed.
Abstract
Periodontitis is a severe gum infection that may result in tooth loss, bone loss and other critical pathological complications. The recruitment of immune cells in the affected area creates a unique microenvironment in which diverse cell types can be found. Recently, a group performed single-cell RNA sequencing (scRNA-seq) to profile the transcriptional landscape of PBMCs of periodontitis. The group identified indicators of inflammatory responses and made suggestions on therapeutic targets. Aligned scRNA-seq data was reported in the Gene Expression Omnibus (GEO) database. In this paper, the GEO data was analyzed and constructed a neural network capable of classifying periodontitis patients using CRIP1 and IFITM1 as the input to the model. The model accurately classified (> 90% accuracy) the test dataset with noise added.
Keywords: Gene expression omnibus database, gene marker, neural network, PBMCs of periodontitis, single cell RNA sequencing