Defining host–pathogen interactions employing an artificial intelligence workflow

Abstract

For image-based infection biology, accurate unbiased quantification of host–pathogen interactions is essential, yet often performed manually or using limited enumeration employing simple image analysis algorithms based on image segmentation. Host protein recruitment to pathogens is often refractory to accurate automated assessment due to its heterogeneous nature. An intuitive intelligent image analysis program to assess host protein recruitment within general cellular pathogen defense is lacking. We present HRMAn (Host Response to Microbe Analysis), an open-source image analysis platform based on machine learning algorithms and deep learning. We show that HRMAn has the capacity to learn phenotypes from the data, without relying on researcher-based assumptions. Using Toxoplasma gondii and Salmonella enterica Typhimurium we demonstrate HRMAn’s capacity to recognize, classify and quantify pathogen killing, replication and cellular defense responses. HRMAn thus presents the only intelligent solution operating at human capacity suitable for both single image and high content image analysis.Editorial note: This article has been through an editorial process in which the authors decide how to respond to the issues raised during peer review. The Reviewing Editor's assessment is that all the issues have been addressed (see decision letter).

Human GBP1 differentially targetsSalmonellaandToxoplasmato license recognition of microbial ligands and caspase-mediated death

“…The cytosolic dye CellMask is excluded from PVs but enters the PV once the PV membrane (PVM) is disrupted by detergent-mediated permeabilization (Figure S1A). Increased dye intensity within Tg vacuoles could be reliably quantified using our artificial intelligence-based high-throughput image analysis workflow HRMAn (Fisch et al, 2019b) (Figure 1A and Figure S1B). These analyses revealed an increase in CellMask staining of PVs in IFNγ-primed THP-1 WT cells, indicating their disruption (Figure S1B).…”

Mimicry embedding for advanced neural network training of 3D biomedical micrographs

“…More often than not host-pathogen biomedical datasets are not large enough for deep learning. However, we reasoned that advances in high-content fluorescence imaging (23) which allow for 3-D, multi-position single-pathogen resolution can serve to increase the size of datasets for ANN analysis (13). To classify single-pathogen data in 3D biomedical images we developed 'ZedMate', an ImageJ-Fiji (24) plugin that uses the Laplacian of Gaussian spot detection engine of TrackMate (25).…”

“…croscopy, detecting and quantifying intracellular viability at the single parasite level is challenging (13). To generate a Tg viability training dataset, cells infected with Tg-EGFP (c1) were fixed and stained with fluorescent markers of DNA (c2), and host cell ubiquitin (c3) which was used a weak label to annotate the subset of "unviable" parasites (13,29) (Fig. 5a).…”

“…Artificial Neural Networks (ANN) excel at a plethora of pattern recognition tasks ranging from natural language processing (1) and facial recognition (2) to self-driving vehicles (3,4). In biology, recent advances in machine learning and Deep Learning (5)(6)(7) are revolutionizing genome sequencing alignment (8), chemical synthesis (9,10) and biomedical image analysis (11)(12)(13). In the field of computer vision, convolutional neural networks (CNNs) perform object detection and image classification at a level matching or surpassing human analysts (14).…”

Deep learning of virus infections reveals mechanics of lytic cells

“…Particularly deep learning methods allow for complex data analysis, often on par with human capabilities (17,18). AI and deep learning are increasingly used in biomedical sciences, for example for image recognition, restorations upon low light sampling, and object segmentation (19)(20)(21)(22)(23)(24)(25). Recently, AI applications for pattern recognition in the life sciences have been immensely enhanced by artificial neural networks (ANNs).…”