Vignettes in R and reference code

I enjoy learning different analyses and interesting ways to show results. I hope this page will serve as a useful reference!

Data wrangling, visualization, and statistics in R:

• tidyverse demo from a fruit fly dataset and features survival analysis and machine learning with SVM
  

• Sample size estimation and stratified block randomization - a demo on estimating sample sizes based and conducting randomization
  

• SQL and R demo- learning how to connect R with a SQL database and building my own SQL database

• Making maps in R and performing PCA on climate rasters and a leaflet demo for making interactive maps

Exploring Kaggle datasets:

Kaggle has a wealth of datasets for data scientists to explore different analyses.

• Breast Cancer survival analysis
  
• Stroke Prediction dataset, exploratory analyses

Decision making under uncertainty:

• Vignette on the general application of Dempster-Shafer Theory for decision making
  
• My shinyapp for binary classification using Dempster-Shafer Theory (DST). This specific use case of DST accounts for model uncertainty for predictions and then combining different sources of evidence.

Simulations: traits as functions

A lot of my PhD was spent thinking about understanding complex characteristics of any organism. Often times, traits can change across an environmental gradient and can be represented as a function. To understand traits as functions, I’ve conducted a few simulations in the area of temperature stress from genes, proteins, to whole organisms.

• Simulating performance curves of genetic clones and understanding potential evolutionary trajectories using a statistical genetics approach

• Simulating the metabolic rates of diapausing animals

• Fitting boltzmann curves to transcript abundance data

• Simulating unfolding parameters and interpreting their biological significance: Unfolding curve parameter tweaks

Unpublished projects:

• Surveyed ants for stress markers (gene expression) that were exposed to experimental warming in nature
  • Script starts from calculating log2 fold change in gene expression to statistical models to predicting stress back out into nature.

Back to main page

---