I am a young computational biologist π©π»βπ», just trying to find her place in the world π
- π Iβm currently working on multiple image-based analysis projects in the Gregory P. Way lab.
- π± Iβm currently learning how to grow my knowledge of Python and R. One of my biggest goals is to improve my biostatistics and machine learning skills.
If you are curious about the projects I am working on, here is a list of the repositories:
- NF1 Cell Painting Project - Image analysis and image-based profiling of Schwann cells of different genotypes (wildtype and null), which is used to generate a machine learning model.
- CFReT Project - Image analysis, image-based profiling, and machine learning to predict healthy or failing cell types of cardiac fibroblasts.
- Pyroptosis Signature Project - Image analysis and image-based profiling of PBMCs and SHSY-5Ys to use in cell death prediction.
- We submitted a preprint for our project titled "Toward generalizable phenotype prediction from single-cell morphology representations". This paper describes our work to make a generalizable model predicting phenotype based on nucleus features utilizing the Mitocheck 2015 labeled dataset. We applied this model to the JUMP consortium and performed a Leave One Image Out analysis, discovering that many features are variable across datasets, leading to problems with generalizability. We propose different steps to solving this issue within our discussion.