The robust response to signaling and developmental cues is driven by singleton and clustered enhancers. The effect of clustered enhancers on target genes is believed to be the sum-of-all enhancers. We tested the relationship among the enhancers within the clusters using estrogen driven enhancer clusters and a dense enhancer cluster at the developmentally regulated INK4/ARF locus. In this talk, I will show the evidences that there exist a complex relationship among the enhancers within clusters. Upon estrogen stimulation, ER binds in the genome in clustered manner where one lead enhancer in the cluster is pre-marked (pre-seeded) by an unliganded receptor even prior to signaling. These clusters emerge upon signaling and though they disappear at the end of signaling, they leave behind the same pre-marked enhancer still bound by the receptor for the next round of ligand exposure. These enhancer clusters are identified as phase separated ER-condensates and any perturbation in pre-seeding abolishes the enhancer-cluster formation and therefore, the transcription of genes and also their periodicity.

Interestingly, though ER enhancer clusters regulate expression of estrogen target genes, they do not exhibit the features of canonical super-enhancers (high levels of Mediator, prescribed distance between enhancers). Furthermore, as opposed to conventional thinking that super enhancers regulate target gene as a sum-of-all-enhancers, ER clustered enhancers exhibited sole dependence on bookmarked enhancers. These observations prompted us to functionally dissect the relationship of individual enhancers in developmentally regulated enhancer cluster. We tested these concepts on one of the most dense-enhancer clusters in the genome at the 9p21 locus which is a GWAS hotspot. We observed that not all enhancers within a cluster are functional and that functional enhancers form an interdependent network where they rely on each other for a net effect on target gene transcription. Furthermore, not all functional enhancers are biochemically marked with high levels of enhancer marks. These observations highlight the need of enhancer cluster dissections in one locus at a time manner to understand the gene regulatory mechanisms driven by enhancer clusters.

Mediator is a multiprotein complex that serves as a connecting link between transcription factor and transcriptional machinery. The subunits are arranged in four different modules; head, middle, tail and kinase. These modules establish contacts with plethora of other proteins to regulate the process of transcription. Our study suggests that presence of intrinsically disorder regions in several subunits makes the Mediator complex very flexible. There is not much study on the holistic view of plant Mediator complex. So, we generated an interaction map of Arabidopsis Mediator subunits to understand the arrangement of the subunits followed by mapping the domains involved in these interactions. Structures of interacting regions were predicted by using bioinformatics tools and then placed together to expound the overall structure. In this study, MED14 and MED17 turned out to be very important for holding the complex. So, we are focussing on these two subunits to understand the functioning of Mediator in plants. Our studies suggest an important role of MED17 in thermomorphogenesis in the model plant Arabidopsis thaliana. I will present data to explain the previously unexplored role of jasmonate signaling, MED17, and their integration with auxin signaling during thermomorphogenesis. Further, our studies reveal critical role of MED17 in transcriptional optimization of root system architecture in Arabidopsis. In genomics studies, we found significant association of polymorphism in MED15 gene with seed size trait in rice. We confirmed the role of MED15 in seed development and deduced the underlying molecular mechanism. Through these examples, I will try to establish the role of Mediator as a processor of transcriptional signals. In all these studies, we have used combinations of bioinformatics and experimental tools.

Regulatory networks are essential for maintaining cellular function, orchestrating the precise expression of genes in response to various internal and external signals. Understanding these networks is crucial for uncovering how cells coordinate complex processes such as development, stress response, and disease progression. The logic of gene regulation involves intricate interactions between transcription factors, enhancers, and other regulatory elements that control gene expression. To capture this logic, we will use co-expression patterns that reveal genes that are regulated together in response to common stimuli. I will briefly introduce RNA sequencing and proteomics which offers a powerful method for measuring gene and protein expression, providing a detailed snapshot of cellular activity. These methods can be applied to bulk samples, individual cells, or tissues in space (spatial transcriptomics), enabling exploration of regulatory networks across biological scales. To identify co-expression networks, mathematical tools like Weighted Correlation Network Analysis (WCGNA) are used to find regulatory modules by analyzing correlations between gene expression patterns, helping uncover key interactions driving gene regulation in health and disease.

Computational biology has revolutionized the study of genome regulation by utilizing various techniques such as RNA-seq, ChIP-seq, Bisulfite-seq, MNase-seq, Hi-C sequencing, and molecular markers. These tools help explore genomic and epigenomic regulation in plants. For example, in our study RNA-seq analysis of transgenic cotton lines revealed that the histone deacetylase GhHDA5 regulates development-related genes crucial for fibre development. ChIP-seq data showed that histone acetylation, particularly H3K9Ac, influences fibre-related genes and their expression during treatment with AA and TSA, highlighting its role in fibre development. Besides, our lab also used bisulfite sequencing to identify differentially methylated regions in Arabidopsis hybrids, revealing that DNA methylation plays a significant role in regulating genes linked to heterosis. Further, to understand the impact of NPR1 on nucleosome remodelling, MNase-seq analysis of NPR1\u2019s role in nucleosome remodelling suggested its involvement in centromeric nucleosome rearrangement.

Our lab also focused on identifying long-range chromatin interactions; Hi-C sequencing was used to capture long-range chromosomal interactions and provide insights into the nuclear organization and functional chromatin architecture. Additionally, molecular markers are crucial for understanding genetic diversity and mapping essential traits, such as productivity and stress tolerance in crop plants. Our GWAS (genome-wide association study) analysis identifies thousands of molecular markers associated with several agronomically essential traits like fibre yield, quality, and plant architecture. Overall, computational biology, functional genomics, and epigenetics are vital for unravelling the complexities of plant traits and can be harnessed to improve crop performance.

Unlike animals, variability in transcription factors (TFs) and their binding regions (TFBRs) across the plants species is a major problem that most of the existing TFBR finding software fail to tackle, rendering them hardly of any use. This limitation has resulted into underdevelopment of plant regulatory research and rampant use of Arabidopsis-like model species, generating misleading results. Here, we report a revolutionary transformers-based deep-learning approach, PTFSpot, which learns from TF structures and their binding regions\u2019 co-variability to bring a universal TF-DNA interaction model to detect TFBR with complete freedom from TF and species-specific models\u2019 limitations. During a series of extensive benchmarking studies over multiple experimentally validated data, it not only outperformed the existing software by >30% lead but also delivered consistently >90% accuracy even for those species and TF families that were never encountered during the model-building process. PTFSpot makes it possible now to accurately annotate TFBRs across any plant genome even in the total lack of any TF information, completely free from the bottlenecks of species and TF-specific models.

Genomic rearrangements (GRs) drive the evolution of genome order. While their role in speciation is well established, the underlying adaptive significance remained underappreciated. Through inference of clade-specific evolutionary GRs in mammals, we observed that the genes flanking the break-points were associated with the clade-specific traits and exhibited greater rates of expression divergence in the clade that had lost the synteny. We present multiple lines of evidence for the chromosomal position effects that largely explain the underlying mechanisms of divergence. We discuss the case of a specific GR present in certain rodent clades and likely confers cancer resistance. We conclude that the clade-specific genomic rearrangements likely modulated the evolution of clade-specific traits and may have a wide-spread adaptive significance.

Bayesian network analysis (BNA) is a powerful probabilistic framework that models complex dependency structures among biological variables using directed acyclic graphs and conditional probability distributions. In bioinformatics, this approach has been successfully applied to reconstruct gene regulatory networks, predict critical interactions, and integrate multi-omics data. Its ability to incorporate prior biological knowledge and handle uncertainty makes it especially suited for high-dimensional, noisy datasets typical of genomics and proteomics studies. In the context of transcription factor (TF) interactions, BNA has been employed to unravel TF–TF networks, revealing combinatorial and cooperative regulatory patterns that drive gene expression programs. By integrating diverse datasets such as ChIP-seq and gene expression profiles, these approaches can distinguish direct from indirect interactions, thereby elucidating the intricate architecture of transcriptional regulation. Similarly, BNA has been adapted to explore miRNA–RNA binding protein (RBP) interactions in miRNA biogenesis, offering insights into the postranscriptional regulation of gene expression. Through the integration different high-throughput data such as CLIP-seq, RNA-seq, and sRNA-seq, Bayesian methods help construct condition-specific causal networks that capture the dynamic interplay between miRNAs and RBPs. Such models have not only improved our understanding of miRNA biogenesis but also enhanced the prediction of miRNA expression levels in various cellular conditions. Overall, Bayesian network analysis provides an integrative and interpretable framework for dissecting both TF–TF and miRNA–RBP interactions. By leveraging heterogeneous datasets, BNA advances our capacity to identify novel regulatory mechanisms, thereby offering promising avenues for deciphering the complex regulatory architectures of biological systems.