Most viral infections result in mitochondrial damage, either because the membranes of this organelle are used for viral replication or because of the type of cellular immune response. 

    This damage can be partially understood by characterizing mitochondrial morphological changes in virus-infected cells; depending on the changes, one can propose whether the immune response against the pathogen is proinflammatory or antiinflammatory by studying thin-section TEM micrographs.

 

     This idea was first studied as part of Dr. Ignacio Lara-Hernandez's Ph.D. thesis on how human Respiratory Syncytial virus affects the mitochondria of cultured cells. However, this study had some limitations, including the relatively small number of mitochondria characterized by thin-section TEM micrographs, the incredibly laborious and time-consuming process of sectioning and measuring this organelle, and possibly observational bias. Nonetheless, this study was fundamental for our group, as it allowed us to correlate the expression of key mRNAs associated with mitochondrial fusion or fission with the ultrastructue of this organelle.

​

       To take Ingacio's project a step forward, we started a very successful collaboration with Prof. Aldo Mejía-Rodríguez, an expert in Deep Learning Image Analysis. First, we developed a methodology with the then Master's student Brianda Alexia Agundis-Tinajero (now a Ph.D. student) to segment and classify mitochondria in cells infected with either SARS-CoV-2 or Zika virus using neural networks. This resulted in a publication in 2024. Now we are taking this model far beyond, so you can feed it micrographs, and it will automatically detect and segment mitochondria in cultured cells and determine whether the organelle is undergoing fusion or fission. Furthermore, it will enable DL-based analysis of mitochondrial damage using Yoshi's classification.

​

​

​

​

​

​

​

​

​

​

​​

Figure taken from: Agundis-Tinajero, B.A.; Coronado-Ipiña, M.Á.; Lara-Hernández, I.; Aparicio-Antonio, R.; Aguirre-Barbosa, A.; Barrera-Badillo, G.; Aréchiga-Ceballos, N.; López-Martínez, I.; Castillo, C.G.; Labrada-Martagón, V.; et al. Deep Learning-Based Automatic Segmentation and Analysis of Mitochondrial Damage by Zika Virus and SARS-CoV-2. Viruses 2025, 17, 1272. https://doi.org/10.3390/v17091272​

​

​

Update. November 2025