Research

Keeping Homeless Youth off Opioids: Optimizing Intervention Delivery using Causal Inference

  • Built a causal inference model which predicts the likelihood of opioid addiction among homeless youth.
  • Minimized the expected number of homeless youth suffering from opioid addiction by assigning people to appropriate interventions using optimization formulation. Paper submitted to AAAI 2020.
  • Github repository: https://github.com/Roopali24/OpioidProject

Assessing the reproducibility of high-throughput experiments in case of missing data.

  • Built a copula-based regression framework to estimate the effect of experimental factors on the reproducibility of results, when there is missing data.
  • Illustrated the usefulness of the method using a study of HCT116 cells from single cell RNA-seq libraries made using microfluidic method and tube-based methods.
  • Further, implemented the method to find the required cost-effective sequencing depth with sufficient reproducibility.
  • Presented at ENAR 2019 and JSM 2019.
  • Github repository: https://github.com/Roopali24/CCR-missing-data