Novel machine learning and data science tools to study infection phenotypes in cells

Advances in Machine Learning and Deep Learning are revolutionising our abilities to analyse biomedical images. These algorithms may allow unambiguous scoring of virus-infected and uninfected cells in the absence of specific labelling. Furthermore, accompanied by interpretability approaches, the ability of convolutional neural networks to learn representations, without explicit feature engineering, may allow for uncovering yet unknown phenotypes in microscopy. In our recent work, we employ the CapsNet architecture equipped with a discriminator and generator. This allowed us to differentiate between intracellular and extracellular Vaccinia virus particles through a classification task using the discriminator part of the architecture. Additionally, using the generator part we were able to visualise the differences between these particles. Finally, we show how a phenotype-centric open-source Python package we developed can facilitate the data science work on virological plaque assay. We designed the package to provide biologists with tools that make phenotypes as intuitive as data frames. We show that our phenotype-centric library can be employed for a range of pathogens like Vaccinia virus or Coronavirus, as well as variations of virological plaque assay including fluorescence-based or crystal violet staining-based assay images.