Sachin Kamboj

Analysis of protein domains and Rett syndrome mutations indicate that multiple regions influence chromatin-binding dynamics of the chromatin-associated protein MECP2 in vivo

The methyl-CpG-binding protein 2 (MECP2) serves both organizational and transcriptional functions in the nucleus, with two well-characterized domains integrally related to these functions. The recognition of methylated CpG dinucleotides is accomplished by the methyl-binding domain (MBD), and the transcriptional repression domain (TRD) facilitates protein-protein interactions with chromatin remodeling proteins. For each known function of MECP2, chromatin binding is a crucial activity. Here, we apply photobleaching strategies within the nucleus using domain-deleted MECP2 proteins as well as naturally occurring point mutations identified in individuals with the neurodevelopmental disorder Rett syndrome (RTT). These studies reveal that MECP2 is transiently associated with chromatin in vivo and confirm a central role for the MBD in directing the protein to heterochromatin. In addition, we report for the first time that the small region between the MBD and the TRD, known as the interdomain region (ID), stabilizes chromatin binding by MECP2 independently of the MBD. The TRD of MECP2 also contributes towards chromatin binding, whereas the N- and C-termini do not. Some common RTT missense and nonsense mutations significantly affect binding kinetics, suggesting that alterations in chromatin binding can result in protein dysfunction and hence a disease phenotype.

Organizational self-design in semi-dynamic environments

In this paper we propose a run-time approach to organization that is contingent on the task structure of the problem being solved and the environmental conditions under which it is being solved. We use TÆMS as the underlying representation for our problems and describe a framework that uses Organizational Self-Design (OSD) to allocate tasks and resources to the agents and coordinate their activities.

Using name-internal and contextual features to classify biological terms

There has been considerable work done recently in recognizing named entities in biomedical text. In this paper, we investigate the named entity classification task, an integral part of the named entity extraction task. We focus on the different sources of information that can be utilized for classification, and note the extent to which they are effective in classification. To classify a name, we consider features that appear within the name as well as nearby phrases. We also develop a new strategy based on the context of occurrence and show that they improve the performance of the classification system. We show how our work relates to previous works on named entity classification in the biological domain as well as to those in generic domains. The experiments were conducted on the GENIA corpus Ver. 3.0 developed at University of Tokyo. We achieve f value of 86 in 10-fold cross validation evaluation on this corpus.

An Investigation of Various Information Sources for Classifying Biological names

The classification task is an integral part of named entity extraction. This task has not received much attention in the biomedical setting, partly due to the fact that protein name recognition has been the focus of the majority of the work in this field. We study this problem and focus on different sources of information that can be utilized for the classification task and investigate the extent of their contributions for classification in this domain. However, while developing a specific algorithm for the classification of the names is not our main focus, we make use of some simple techniques to investigate different sources of information and verify our intuitions about their usefulness.

A comparison of two GIV mechanisms for providing ancillary services at the University of Delaware

At the University of Delaware, we are providing ancillary services by controlling the bidirectional power transfer between 15 EVs and the grid. To control this power transfer, a set of algorithms, models and interactions is used, called a “GIV (Grid Integrated Vehicle) mechanism”. In literature, many GIV mechanisms are proposed. However, because these mechanisms are evaluated independently in specific scenarios, their differences are not always clear. In this paper, we take a first step in tackling this challenge by comparing two different GIV mechanisms in the same scenario at the University of Delaware: a decentralized and a centralized mechanism. In the decentralized mechanism, which is currently operational at our test environment, EVs decide autonomously on the amount of power available for ancillary services. In the centralized mechanism, a central server gathers all EV information and makes a decision for all EVs. In evaluation, both GIV mechanisms are compared with each other. Simulation results show that the centralized mechanism outperforms its decentralized counterpart in terms of available power for ancillary services. On the other hand, the decentralized mechanism enables large-scale integration by distributing computations across all EVs.

Exploring Robustness in the Context of Organizational Self-design

Robustness is the ability of a multiagent system to recover from failures and exceptions. In particular, the system should be able to recover from task and agent failures and the failure of any single agent or group of agents should allow the graceful, predictable degradation of performance. Hence, designing robust systems is a critical challenge facing many multiagent designers. Organizational Self-Design (OSD) has been proposed as an approach to constructing suitable organizations at runtime in which the agents are responsible for constructing their own organizational structures. OSD has also been shown to be especially suited for environments that are dynamic and semi-dynamic. However, the problem of making these self-designed organizations robust is still an open research problem that has not been studied to any considerable extent. In this paper, we focus on developing and evaluating robustness mechanisms that can be used by the agents in conjunction with OSD.

Vehicle to Grid Demonstration Project

This report summarizes the activities and accomplishments of a two-year DOE-funded project on Grid-Integrated Vehicles (GIV) with vehicle to grid power (V2G). The project included several research and development components: an analysis of US driving patterns; an analysis of the market for EVs and V2G-capable EVs; development and testing of GIV components (in-car and in-EVSE); interconnect law and policy; and development and filing of patents. In addition, development activities included GIV manufacturing and licensing of technologies developed under this grant. Also, five vehicles were built and deployed, four for the fleet of the State of Delaware, plus one for the University of Delaware fleet.