Back

Deep Learning-Driven Discovery of Mitochondrial Factors Modulating Influenza A Virus Infection

Shkel, O.; Kim, H.; Haque, M. M.; Kharkivska, Y.; Hong, K. T.; Lee, S.-H.; Kim, Y. K.; Song, C.-S.; Kim, W. Y.; Lee, J.-S.

2026-02-26 microbiology
10.64898/2026.02.25.707858 bioRxiv
Show abstract

Influenza virus remains a major public health threat, highlighting the need to identify host proteins that regulate viral replication. In this study, we identify several mitochondrial proteins that influence virus-cell interactions using a Convolutional Neural Network (CNN) and a proximity labeling approach. Using CNN, we analyzed cell morphology and morphological changes of nucleus, mitochondria, and endoplasmic reticulum before and after influenza infection. Among these, mitochondrial morphology provided the clearest separation between infected and uninfected cells, achieving the highest classification precision of 84.9%. To uncover mitochondrial factors involved in infection, we performed APEX2-based proximity labeling mass spectrometry of the mitochondrial proteome. Functional validation revealed that knockdown of CH60, ETHE1, and SQOR increased IFN-{beta} mRNA levels, while knockdown of LONM enhanced influenza vRNAs accumulation. Moreover, depletion of CH60, ETHE1, LONM, MPPB, and SQOR significantly altered production of progeny virus. Together, these findings demonstrate that several mitochondrial matrix and inner membrane proteins can impact influenza virus replication within host cells.

Matching journals

The top 6 journals account for 50% of the predicted probability mass.

50% of probability mass above

"Similar papers" are the closest papers from that journal in the model's embedding space. They show what the match is built on, but the ranking comes mostly from a classifier over the whole training set, not from these examples alone.