Selection and Transmission of Antimicrobial Resistance in Complex Systems (STARCS)




Antibiotic-resistant bacteria are not new, but in the last two to three decades antibiotic resistant bacteria have been constantly increasing and rapidly spreading. As a result, the emergence of antibiotic resistance is increasingly limiting treatment options for mild to moderate infections, which makes it of pivotal importance to control the emergence and spread of antibiotic resistant nacterial pathogens.

Ongoing project

In the STARCS project we study the pathways by which resistant bacteria, particularly enterococci resistant to vancomycin and enterobacteriales resistant to extended-spectrum beta-lactamases, can emerge in complex ecosystems, such as the human intestinal tract, in which they can acquire genetic material that encodes resistance to antibiotics.

To this aim, we have developed a number of tools that are used to characterise the reservoir of resistant bacteria and resistance genes in microbial ecosystems. In addition, we perform research into the transfer of resistant bacteria and resistance genes between animals and humans. As antibiotic resistance can spread via horizontal gene transfer of mobile genetic elemenst (MGE) in addition to the spread of antibiotic resistant strains, we have developed in the STARCS consortium novel tools to identify and reconstruct MGE’s based on whole genome sequence data and to characterize with great precision the MGE-bacteria interaction networks of the microbiota in the mammalian gut using an approach called metaHi-C.

Using all these tools studies comparing bacteria found in humans and farm animals in the Netherlands by analysing their DNA revealed that resistant bacteria found in humans were different from those found in farm animals. This shows that humans are rarely infected by antibiotic resistant bacteria that come directly from farm animals.

Furthermore, three primary studies conducted in a community setting and two in hospital settings, using samples collected from different sources: patients, live and slaughtered food animals, and environmental samples in the slaughterhouses, taken from developing (Vietnam) and developed countries (Italy, Germany) with different antibiotic use and antibiotic resistance backgrounds, revealed, again, limited contribution of food-animal sources, particularly chickens and pigs, to causing urinary tract infections by ESBL-resistant enterobacteriaceae (ESBL-E) in Hanoi, Vietnam.

In hospital settings, actively screening and decolonising patients carrying ESBL-E are promising infection control strategies. Using simple mathematical models, we were able to show that the environment that is shared between farm animals and humans might be an important pathway for the transmission of antibiotic resistant bacteria as it can act as a reservoir between the two populations. Because of this, measures such as curtailing antibiotics in farm animals have much less impact on human health.

Furthermore, our modelling showed that the effect of curtailing of antibiotic resistance in farm-animals on human foodborne diseases is two-fold. By curtailing antibiotics is farm animals the proportion of human foodborne illnesses that is caused by a resistant pathogen will be decreased, but the number of food-borne illness cases is likely to increase due to this measure. However, further modelling showed that this can be mitigated by maintaining good farm bio-security (farm-to-fork and livestock health) and reducing transmission from animals to humans.

We furthermore looked at how (international) food imports affect the efficacy of antibiotic curtailment in livestock. For this we used a simple mathematical model. This shows that imports from non-domestic sources can reduce the efficacy of local livestock antibiotic curtailment.

Project partners

  • Rob Willems, University Medical Center Utrecht, Netherlands (Coordinator)
  • Dik Mevius, Wageningen University & Research, Netherlands
  • Dan Andersson, Uppsala University, Sweden
  • Teresa M Coque, Ramón y Cajal University Hospital, Spain
  • Romain Koszul, Pasteur Institute, France
  • Mark Woolhouse, University of Edinburgh, United Kingdom
  • Surbhi Malhotra-Kumar, University of Antwerp, Belgium 

Project resources


  • mlplasmids: Consists of binary classifiers to predict contigs either as plasmid-derived or chromosome-derived.
  • gplas: A tool to bin plasmid-predicted contigs based on sequence composition, coverage and assembly graph information.
  • RFPlasmid: Predicting plasmid contigs from assemblies.
  • ResCap: Repository for software, raw data tables and data bases.
  • PATO: A R package designed to analyze pangenomes (set of genomes) intra or inter species.
  • MetaTOR: Metagenomic Tridimensional Organisation-based Reassembly – A set of scripts that streamlines the processing and binning of metagenomic metaHiC datasets.