Seminars

This is a hybrid talk and zoom link to attend is: https://zoom.us/j/99127472570 Meeting ID: 991 2747 2570 Passcode: 217863 ABSTRACT Humans and ecosystems are frequently exposed to myriad of chemicals, including chemicals in consumer products, industrial pollutants, pesticides, among others, which collectively constitute the chemical exposome. These chemicals can persist in the environment and bioaccumulate, leading to detrimental effects on humans and other organisms, as well as long-term ecological impacts. Therefore, it is imperative to characterize the chemical exposome and its impact on human and ecosystem health. To this end, traditional toxicity testing often relies on animal models which can be low-throughput, expensive and time consuming, and therefore, computational approaches have emerged as effective alternatives to expedite the characterization of the ever-expanding chemical exposome. In this thesis, we employ different computational approaches to investigate the structure-activity landscape of environmental chemicals in the exposome, and further, integrate heterogeneous toxicological datasets to elucidate chemical-induced adverse effects on both human and ecosystem. For the analysis of the structure-activity landscape of endocrine disruptors, we focus on two distinct environmental chemical spaces, namely, androgen receptor (AR) binding chemicals and thyroid stimulating hormone receptor (TSHR) binding chemicals. For both chemical spaces, we employ several computational approaches to analyze the presence of heterogeneity in the structure-activity landscape of these chemical spaces and identify the presence of activity cliffs, i.e., structurally similar chemicals exhibiting large differences in their activities against a target receptor. Further we classify the identified activity cliffs based on the information of their chemical structures. Additionally, we analyze the structure-mechanism relationships of the TSHR binding chemicals and identify structurally similar chemicals differing in their mechanism of actions. In sum, the inferences from these computational analyses will aid in development of improved toxicity predictors for characterization of the chemical exposome. In addition, to investigate the adverse health effects induced by environmental chemicals, we focus on certain chemical classes namely, inorganic cadmium compounds (heavy metal), plastic additives and petroleum hydrocarbons (PHs). For each case, we curate a list of chemicals by relying on published reports and existing resources. We then integrate the biological endpoint data from various toxicological resources to identify associations between the chemicals and the high confidence adverse outcome pathways (AOPs) within AOP-Wiki. Further we construct chemical-specific AOP networks and analyze toxicity pathways to understand the mechanisms underlying chemical-induced adverse effects in both humans and aquatic organisms. Additionally, we assess the toxicities of the PHs across diverse ecological species using network-based approaches and perform an ecological risk assessment for a group of PHs. In conclusion, this thesis presents a systematic computational investigation of environmental chemicals and their adverse health effects on humans and ecosystems, thereby providing a holistic overview of chemical exposome and its implications on health from a One Health perspective.

Dynamical generation of mass is of fundamental importance in many areas of physics all the way from strongly interacting electrons in condensed matter to quantum chromodynamics. I will mainly focus on three (Euclidean) dimensional theories and show how lattice regularization can help probe scale symmetry. If time permits, I will also discuss two-dimensional QCD.

Classical continued fraction (CF) theory represents numbers as, e.g., pi=3+1/(7+1/(15...)), and has applications throughout number theory, geometry, and dynamics. While real CFs are well-studied, little is known about complex CFs and other higher-dimensional CF algorithms. In this talk, I will start by providing some intuition for the geometry and dynamics of CFs using some interactive visualizations developed by undergraduate students; and then will share new results on convergence, mixing properties, and invariant measures for CF algorithms in R^n and its non-Euclidean analog the Heisenberg group.

Celestial holography is the study of a putative duality between 4D asymptotically flat spacetimes and a 2D "celestial" conformal field theory living on the boundary two-sphere. It is interesting to study marginal deformations of celestial CFTs which are the result of adding in a marginal, dimension Delta=2, operator into the spectrum. In this talk, I will cover the basics of celestial holography including the definitions of operators on the boundary, their relation to bulk fields and how to compute the respective OPE coefficients. Starting with Yang-Mills in the bulk, I will describe how to construct marginal operators on the boundary and then discuss how we can use them to learn about the geometry of the associated conformal manifold.

In holography, the bulk reconstruction program has its origin in the context of AdS/CFT correspondence. While the AdS/CFT correspondence maps bulk fields to boundary operators. It has also been extended to heavy scalar fields (i.e fields in the principal series representation)in de Sitter space time. In this talk I will carry out bulk reconstruction obtain boundary representations for scalars of all masses as well as higher spin fields. this extrapolated map also extended the construction from Bunch-Davies to α vacuum.

In this talk, we introduce a new class of diagram algebra: the walled Klein-4 Brauer algebra. This algebra is explored through its generators and relations, and we detail the indexing set of its simple modules. Additionally, we establish the Robinson-Schensted correspondence for the walled Klein-4 Brauer algebra, constructing a bijection between the set of all walled Klein-4 Brauer diagrams and pairs of Klein-4 vacillating tableaux.

It is a learning seminar.

From coherent swarms of fish or movement of ions into cell membranes to leakage of gases through container pores or waves propagating through biological tissues, the non-equilibrium world around us is complex, diverse and dynamic. In this talk, I will first briefly discuss a few classes of non-equilibrium theories of interacting systems which comprise rich phase behaviour and lend themselves to analytical tractability. Next, motivated by contraction waves in biological tissues, I will introduce a new minimal hydrodynamic description of a class of pulsating systems and discuss transitions between different dynamical phases therein. I will then show how fluctuations affect such pattern forming systems. I will conclude by sharing broader perspectives on how deformable systems can provide a fertile ground for Statistical Physics research in the near future.

We present deterministic suffix-reading automata (DSA), a new automaton model over finite words. Our motivation for this model comes from specifications which can be described using rules of the form: "when a certain sequence of inputs is seen, perform this action". DSAs provide a cleaner representation, compared to deterministic finite automata (DFA), in such situations. Transitions of a DSA are labeled with words. From a state, a DSA triggers an outgoing transition when a word ending with the transition-label is seen. Therefore, rather than moving along an input word letter by letter, a DSA can jump along blocks of letters, with each block ending in a suitable suffix. This feature allows DSAs to recognize regular languages more concisely. In this talk, we will focus on questions around finding a minimal DSA for a regular language. A key technical ingredient that we will describe is a "natural" method to derive DSAs from a given DFA: a DFA-to-DSA conversion. We make a (surprising) observation that the smallest DSA that we derive from the canonical DFA of a regular language L need not be a minimal DSA for L. This showcases some fundamental bottlenecks (aka. interesting challenges) in determining minimal DSAs. In fact, we prove that given a DFA and a number k, deciding whether there exists a DSA with size less than k is NP-complete. This is a very fresh model, with plenty of unsolved automata-theoretic and algorithmic questions. We will discuss some of them towards the end of the talk. Joint work with R Keerthan, R Venkatesh and Sagar Verma.