Caracterização e análise de regiões celulares em imagens de marcação fluorescente de larvas da Drosophila melanogaster

Abstract

Drosophila melanogaster, commonly known as fruit fly, is considered one of the main study models used in biomedicine. Its gene expression patterns are widely studied in the form of digital images, allowing for investigation of a wide range of biological phenomena. In this project, images of larvae from this fly are used to identify the role of Calpain A protein in the innate immune response of Drosophila. Using image segmentation techniques on the provided image set, masks were generated for nuclear and cytoplasmic quantification in the context of the fat body (tissue responsible for the production of antimicrobial and antifungal peptides in response to Toll pathway activation). Finally, new visualization tool was implemented using the D3.js library for integrated visualization of segmentation results and main features of nuclear and cytoplasmic regions. As a result, nuclear segmentation was achieved, obtaining masks that allowed for accurate identification of cellular nuclei and quantification of protein expression intensity in the nuclei.