Arup Kumar Ghosh

Decision procedure based discovery of rare behaviors in Stochastic Differential Equation models of biological systems

Stochastic Differential Equation (SDE) models are often used to model the dynamics of complex biological systems. The stochastic nature of these models means that some behaviors are more likely than others. It is often the case that a models primary purpose is to study rare but interesting or important behaviors, such as the formation of a tumor, or the failure of a cyber-physical system. Unfortunately, due to the limited availability of analytic methods for SDEs, stochastic simulations are the most common means for estimating (or bounding) the probability of rare behaviors. Naturally, the cost of stochastic simulations increases with the rarity of the behavior under consideration. To address this problem, we introduce a new algorithm, RESERCHE, that is specifically designed to quantify the likelihood of rare but interesting behaviors in SDE models. Our approach relies on the use of temporal logics for specifying rare behaviors of possible interest, and on the ability of bit-vector decision procedures to reason exhaustively about fixed precision arithmetic. We also compute the probability of an observed behavior under the assumption of Gaussian noise.

A Stakeholders' Analysis of the Systems that Support Foster Care

This paper focuses on understanding how the key stakeholders of the foster care system work together, as well as the systems that facilitate collaboration. We conducted 20 interviews with foster parents, 2 with non-profit agencies, and 1 case worker interview. Our findings suggest that each stakeholder faces their own unique challenges and sociotechnical systems put in place to assist stakeholders are not sufficient for addressing these concerns. This poster identifies the major stakeholders, their goals and needs, as well as the systems used. Our work helps identify gaps in order to design more effective systems that can better support the foster care system.

Carebit: A Privacy-Preserving "Step" Toward Remote Informal Caregiving

Several tele-monitoring systems have been developed for in-home patient use. Unfortunately, many of these systems are cost prohibitive and privacy invasive to the patient. To overcome this problem, we designed a more affordable and lightweight solution called Carebit, an Android application that leverages the Fitbit API. We conducted two user studies to understand ways to improve our design. Overall, we found that the notifications feature is the most useful feature for users, and no concerns about privacy were mentioned. The goal of Carebit is improve informal caregiving.

Understanding User Reviews of Adolescent Mobile Safety Apps: A Thematic Analysis

With the growing use of mobile smart phones among teens, adolescent online safety is becoming more and more challenging. To overcome this problem, parental control applications have been developed. Yet, no one knows why these apps have very low adoption rates nor if they are effective. To address this problem, we previously conducted a structured analysis of existing adolescent online safety apps. In this paper, we briefly summarize our previous results and introduce our new approach for gaining additional insights from the actual users of these apps. We summarize our methodology for doing this and present the results of an initial thematic analysis of user reviews of adolescent online safety apps.

Discovering rare behaviours in stochastic differential equations using decision procedures: applications to a minimal cell cycle model

Stochastic Differential Equation (SDE) models are used to describe the dynamics of complex systems with inherent randomness. The primary purpose of these models is to study rare but interesting or important behaviours, such as the formation of a tumour. Stochastic simulations are the most common means for estimating (or bounding) the probability of rare behaviours, but the cost of simulations increases with the rarity of events. To address this problem, we introduce a new algorithm specifically designed to quantify the likelihood of rare behaviours in SDE models. Our approach relies on temporal logics for specifying rare behaviours of interest, and on the ability of bit-vector decision procedures to reason exhaustively about fixed-precision arithmetic. We apply our algorithm to a minimal parameterised model of the cell cycle, and take Brownian noise into account while investigating the likelihood of irregularities in cell size and time between cell divisions.

A computational metabolic model of the NG108-15 cell for high content drug screening with electrophysiological readout

Computational Systems Modeling could play a significant role in improving and speeding up of the drug development process. By the incorporation of cellular modeling into a High Information Content Drug Screening platform the information content of the pharmacological test could be significantly increased through a deeper understanding of cellular pathways and signaling mechanisms. Unfortunately, many of the cellular signaling pathways in the cells are yet to be explored. Moreover, which is an even larger problem, their integration into a functional signaling network at the whole cell level is almost unknown or untested. Thus, there is an urgent need to develop a data-driven functional whole-cell model which enables the correlation of biochemical and physiological experimental results at the whole cell level with partial information available for the metabolic and signal transduction pathways of the cell. We have built a whole-cell model of NG108-15 cells and validated some of the underlying cellular metabolic and signal transduction networks with a series of detailed experiments in order to predict cellular responses to a wide variety of extracellular stimuli. This validated assay system will be an important tool for the identification of cellular changes and activation of signal transduction pathways based on changes of electrophysiological properties and responses of the cell and would have a high impact on drug screening and toxicity evaluation at the cell-system level.