Computer Vision for Bioinformatics 

This project focuses on developing computational methods to model actin filaments in low-resolution cryo-electron tomography (cryo-ET) density images. Actin filaments play a critical role in cellular processes such as shape maintenance and movement. While cryo-ET can visualize 3D cellular structures, it often produces noisy and anisotropic density maps due to resolution limitations. The goal of this project is to enhance the visibility, segmentation, and reconstruction of actin filaments by integrating advanced image processing techniques and machine learning algorithms, thereby improving the precision of actin analysis and advancing our understanding of cellular dynamics.

This project focuses on the segmentation and quality assessment of protein secondary structures in cryo-electron microscopy (cryo-EM) images. Accurate identification of protein secondary structures, such as alpha helices and beta sheets, is crucial for understanding protein function, interactions, and mechanisms. While cryo-EM enables high-resolution visualization of protein structures, challenges remain in reliably detecting these structures due to factors like noise, resolution limitations, and sample heterogeneity. This project develops advanced computational methods to segment protein secondary structures from cryo-EM density maps and implements quality assessment metrics to ensure precise analysis, ultimately advancing our understanding of protein functions and supporting research in drug design and disease mechanisms.