We are a system virology lab employing a combination of interaction proteomics and functional genomics to understand the bacterial immune response to phage infection and the phage mechanisms responsible for immune evasion.


Bacterial viruses as next-generation therapeutics

Much like human cells, bacteria are susceptible to fatal viral infections. These bacterial viruses (bacteriophages, i.e phages) have gained significant attention in recent years as one of the most promising alternatives to antibiotics to tackle the emerging problem of drug-resistant bacteria.
However, just like our cells have immune systems aimed at eliminating viral infections, bacteria possess powerful anti-phage mechanisms (like CRISPR) which greatly reduce the efficacy of therapy using phages. Likewise, phages evolved mechanisms to escape these systems resulting in the numerous anti-defense systems widespread across phage families.
Our lab’s main goal is to discover these bacterial defense systems, understand their composition and triggers as well as identifying phage mechanisms to evade them.


To study these systems we developed new interaction proteomics techniques for high-throughput generation of protein-protein interaction networks in phage-infected bacteria.

Co-fractionation mass-spectrometry

Most proteins are organized in macromolecular assemblies (protein complexes), which represent key functional units regulating and catalyzing the majority of cellular processes in health and disease. Disregulation of protein complexes by pathogen proteins (either by direct binding or pathway-level perturbation) is a critical determinant of infection outcome.
To identify protein complexes in a high-throughput manner, we utilize co-fractionation mass spectrometry, a novel interaction proteomics technique where the complexes derived from a native lysate are size-fractionated into a large-number of fractions and each fraction is subsequently acquired using highly quantitative mass-spectrometry (DIA-MS). The resulting protein intensity profiles are then used as proxy for assembly state, under the assumption that proteins sharing similar profiles were physically associated at the separation stage. We introduced the first machine learning model for de-novo prediction of protein complexes from co-elution data and continue to develop computational tools for high-throughput generation of interaction networks at scale.

Image1 Schematic of a co-fractionation mass spectrometry experiment

Interested? Reach out for an opportunity to join our team!