Traditional disciplinary boundaries

Integrating computational sciences with machine learning

We bring innovation across the boundaries of traditionally-defined disciplines, and combine theoretical models with data-driven machine learning approaches. Specific applications range from biomarker discovery to novel materials and financial forecasting.

Explore research areas »

Our recent developments

Biomarker discovery from high-throughput data: Variable selection methodology inspired by computational physics techniques (see toolbox overview).

Recent insights by tag (scroll to zoom)

Anything in here will be replaced on browsers that support the canvas element

Research highlights

  • Proteomic biomarkers of breast cancer prognosis: Emphasis on accurately identifying high metastatic risk patients
  • Biocompatibility: Predicting protein resistance properties of molecular monolayers
  • Drug discovery: Accounting water effects when computing small molecule drug-target interactions
  • Water in protein and membrane pores: A coarse-grained molecular theory approach
  • Nanoscale forces: Computing nanoparticle interactions

Physics-inspired biomarker discovery toolbox

Multi-disciplinary Insights has developed biomarker discovery tools that identify the smallest number of key variables most essential for a given biological context from high-throughput datasets. The algorithm is inspired by statistical physics of strongly interacting systems, and uncovers most sensitive predictor combinations for medical diagnosis and prognosis applications.

Read more »

Science mapping from a bibliometric perspective

Network visualization helps uncover interesting and evolving relationships between various areas of scholarly activity. Here we use a sigma.js javascript framework to render an interactive map of science that is based on journal subject categories and aggregated citation data.

Read more »

Science map: Subject categories of scholarly journals

About the map of science dynamic banner

The top banner on this page presents part of the map of science visualization (move the mouse to thetop of the page to explore). This particular network of scientific fields is constructed based on a bibliographic analysis of journal citations:

  • Network nodes correspond to scholarly journal subject categories by Thomson Reuters;
  • Edges are based on aggregated journal-journal citations.

This graph is a good illustration for the interconnectedness of various disciplines and their relationship to each other.

More details can be found on our structure of science page.