The reliability of observational approaches for detecting interspecific parasite interactions: comparison with experimental results

- We assess various methods to detect interactions between coinfecting parasites.
- Correlations are very unreliable, often predicting incorrect associations.
- Longitudinal analyses, capturing the sequence of infection events, perform best.
- The most reliable way to detect interactions may be through perturbation experiments.
- Data from drug treatment programs could be ideal for detecting parasite interactions.