Email: [email protected]
Advisor: [Olga Vitek] (https://www.khoury.northeastern.edu/people/olga-vitek/)
Venue: IEEE Transactions on Big Data
Impact factor: 5.67 (2018)
Additional Information: The paper was published on [IEEE Transactions on Big Data] (https://ieeexplore.ieee.org/xpl/RecentIssue.jsp?punumber=6687317). IEE publishes peer-reviewed scientific articles that work with big data. It is considered as a top journal in intersection of machine learning and biology field.
Contribution statement from Advisor:
Sara's recent paper “Leveraging structured biological knowledge for counterfactual inference: a case study of viral pathogenesis” was accepted to IEEE Transactions on Big Data, in a special issue related to COVID-19 on Dec 14, 2020.
The manuscript contains three main algorithms. The first algorithm (Query2BEL) takes as input a knowledge graph and a causal query and converts it to a biological expression language (BEL). The second algorithm (BEL2SCM) takes BEL as input and converts it to a Structural Causal Model (SCM). The third algorithm (GetCE) estimates the causal effect on a specific variable upon intervention on a set of variables. Sara designed the second and third algorithm, and contributed in implementing both of them. She designed and implemented the first case study (IGF) by creating observational data for the IGF model based on solving the ordinary and stochastic differential equations, designing and implementing an interventional system that takes as input any desired intervention and outputs interventional data, and calculating the causal effect of interest from these data sets based on GetCE algorithm. She came up with a solution on how to create simulation data for the second case study, such that the values of variables stay within the desired range and implemented that. Sara contributed in writing the paper, including the background, methods, the case studies, and the results section. She showed her leadership abilities in this project by leading two master students.