Cell states

Milestones

• Researched current literature to review state of the art methods to meet customer needs.
• Designed and implemented feature engineering and generation pipeline in Wolfram Language.
• Dimension reduction and clustering algorithms to discover emergent cell states.

Summary

Single cells take on many different shapes depending on their specific phenotype. When cells are genetically engineered their overall morphology (shape) can change. The specific aspects of the shape change (i.e. elongation, roughness) can be attributed to more deep rooted biological processes but this can be difficult to untangle. During this project we integrated a feature engineering pipeline into an unsupervised machine learning model to help the client discover emergent cell types.