Big data analysis is key to identify the critical signals and build predictive quantitative models. Sometimes available methods fall short and we end up developing new statistical methods.
All models we build should tell us something that we did not know before. This can send us into unexpected directions, such as enabling us to predict the responses of patient samples to different signals.
Using Pharmacoscopy we can test hundreds of drugs on small blood or tissue samples from individual people, and actually see what each drug does to each individual donor cell, be it healthy or cancerous.
The single-cell resolution allows to compare drug killing of cancer cells with that of healthy donor cells, and to measure the drug-effects on immune-mediated cancer cell killing, ensuring we identify potent treatments.
Pharmacoscopy uses the same markers that clinicians use to diagnose different cancers, and it is compatible with blood samples, bone-marrow biopsies, and excised lymph nodes, making the technology widely applicable.
Unlike many competing methods to identify optimal therapies for individual patients, Pharmacoscopy produces results within days from getting the sample, meaning we can be helpful also in accute circumstances.
From small patient and donor biopsies Pharmacoscopy can measure thousands of different conditions, allowing us to test many drugs in repeat tests and at different concentrations.
Image-based screening in patient biopsies ("pharmacoscopy") identifies effective treatments for patients with blood cancers.Snijder & Vladimer, et al., Lancet Haematology, 2017. - Commentary by JP Bourquin.
We find that image-based screening in blood ("pharmacoscopy") can quantify how drugs alter the immune system.Vladimer & Snijder, et al., Nature Chem Bio, 2017.
We uncovered a remarkable feature in the coregulatory network of hundreds of different membrane lipid species.Köberlin & Snijder, et al., Cell, 2015. -
Here we show that Solute Carriers are the most unevenly studied gene family, despite their importance for diseases.César Razquin & Snijder, et al., Cell, 2015.
We discuss the pros and cons regarding inference of genetic interactions using different techniques and model systems.Liberali, Snijder, and Pelkmans, Nature Reviews Genetics, 2015.
We generated the first functional genetic interaction map of genes regulating 13 different endocytic pathways.Liberali, Snijder, and Pelkmans, Cell, 2014.
Here we discuss solutions to the obstacles standing in the way of efficient data sharing between researchers.Snijder et al., Nature Biotechnology, 2014.
We discovered a new algorithm that accurately infers functional interactions between genes from genome-wide data.Snijder et al., Nature Methods, 2013. - Check a human network.
We created the first classification of 17 mammalian viruses based on the host genes that are involved.Snijder et al., Molecular Systems Biology, 2012.
Using large-scale functional genetic screens we discovered essential proteins involved in the early infection of Vaccinia Virus.Mercer & Snijder et al., Cell Reports, 2012.
Read our discussion on the balance between randomness and order in what is called cell-to-cell variability.Snijder & Pelkmans, Nature Reviews Molecular Cell Biology, 2011.
Using machine learning, we discovered that influences from the cellular environment are a big reason why cells are differentSnijder et al., Nature, 2009. -
We are a publicly funded laboratory headed by Prof. Dr. Berend Snijder at the Institute of Molecular Systems Biology of the ETH Zurich in Switzerland. We are Systems Biologists, which in practice means that we combine large-scale cell biological experimentation with big data analysis and computational modeling. We do this to improve our ability to measure key aspects of cell biology, and to improve our understanding of life in health and disease.
I'm glad you asked. We enjoy several great web comics that are all "mostly" safe for work: