Following are some projects I have worked on that have gained recognition from global experts as significant contributions to the field.
-
Quality control protocol for novel markers for early cancer detection (Lead researcher). I developed a protocol for monitoring haemolysis, addressing issues that can lead to false results in cancer detection. This protocol was published in an international journal and has been cited in over 200 studies, demonstrating its widespread adoption and impact in the field. [Link to the scholarly article]
Methods. R, linear regression, Receiver Operator Characteristic (ROC) curve analysis, optimisation of sensitivity and specificity point for clinical diagnostics.
-
Identifying novel predictors to optimise surgery for patients with ovarian cancer (Lead researcher). Surgical removal of cancer is routinely used in the management of the disease. This study identified novel biomarkers associated with positive surgical outcomes, optimising risk the risk-to-reward ratio for a patient.<a href="https://doi.org/10.1016/j.ygyno.2017.11.005",target="_blank">[Link to the scholarly article]
Methods. R, linear regression, model/feature selection, k-fold cross validation, bootstrapping, support vector machine (SVM) and Diagonal Linear Discriminant Analysis (DLDA).
-
Genetic analysis of breast cancer (Lead researcher). Intron retention, a rare form of alternative splicing, was found to be abnormal in breast cancer. Using publicly available data on more than 1000 patients, this study pinpointed the abnormality in tissue composition and cellular proliferation rate. We discovered a link between proliferation rate and intron retention. [Link to the scholarly article]
Methods. R, Unix, HPC/cloud computing, bootstrapping, RNA-seq, Gene ontology, gene set enrichment analysis (GSEA), hierarchical clustering, K-means clustering, survival analysis and dimension reductional using PCA.
-
Genetic analysis of blood cancer. This study discovered distinct epigenetic marks associated with intron retention in blood cancer. [Link to the scholarly article]
Methods. R, HPC/cloud computing, bootstrapping.
- Identifying High-Performance Stocks Through Fundamental Financial Ratio Analysis. Key financial ratios associated with historic growth rates of stocks were identified using linear and logistic regression, feature selection and PCA. [github link].
Methods. Python, R, unix, API requests, SQL.
- Analysis of the Melbourne Property Market. The goal of this project was to aid the decision-making process of buying a property, including the choice of dwelling type and the effect of suburb demography. This project revealed differences in appreciation rates for apartments vs detached houses in a given radius of a location, ie Melbourne CBD. The results were overlayed on a map for visualisation. PCA was performed on the Australian Bureau of Statistics (ABS) data to identify suburbs that are visually similar in demographic composition. [github link]
Methods. Python, geopanda, seaborn, json, scikit-learn for regression, jupyter notebook, mapbox third-party package.
- Depreciation rates for cars in the USA. This project aimed to identify cars that depreciate at a slower rate in the USA from the data collected from a car seller's website. [github link]
Methods. R, web scraping, data wrangling using regex, linear regression.
- Graduate Diploma in Statistics (2022 - current)
- PhD
- B.Tech (Hons I)