Biology subjects

Ranjan, S.

Publications and source records attributed to Ranjan, S..

11 recordsLinked to original sources

bioRxiv2026

Hepatic Differentiation of Human Pluripotent Stem Cells into Functional In Vitro Models Recapitulating Native Liver Complexity for MASLD Modelling

Human in vitro hepatic models that accurately recapitulate liver function are essential for fundamental and translational research; however, currently utilised models for disease modelling and drug discovery lack physiological fidelity and require prolonged culture time. Here, we present a streamlined 10-day protocol for efficient and reproducible differentiation of human pluripotent stem cells into hepatocyte like cells (HLCs) and hepatic liver organoids (HLOs). Both models exhibited mature hepatocyte differentiation, as evidenced by albumin secretion and CYP3A4 metabolic activity. Interestingly, HLOs display enhanced multicellular complexity incorporating endothelial, stellate, and macrophage populations along with hepatocytes, thereby more closely recapitulating the native liver microenvironment than HLCs. Here, steatosis was induced in both platforms, which resulted in triglyceride accumulation and upregulated lipogenic markers (DGAT1, DGAT2). However, only HLOs recapitulated advanced disease characteristics, including inflammatory (IL-10) and fibrotic (alpha SMA, COL1A1) responses. Resmetirom, a thyroid hormone receptor-{beta} agonist, significantly reduced steatosis and restored molecular signatures in both models. Additionally, transplanted HLOs demonstrated prolonged survival and displayed host-derived vascularization, thereby validating in vivo maturation. Collectively, HLCs provides a rapid and physiologically relevant liver model, with HLOs offering superior utility for disease modelling, therapeutic evaluation, and regenerative applications due to their enhanced functional and physiological relevance.

cell biology↗

bioRxiv2026

WildAlert: A Real-Time, AI-Driven Early Warning System for Wildlife Health and Ecological Threat Detection

Emerging infections and environmental disruptions increasingly threaten wildlife and ecosystem health. Free-ranging wildlife often serve as early indicators of ecological instability, making timely detection of morbidity and mortality events critical for early warning. Yet, existing systems lack the analytical capacity for real-time outbreak detection. We present WildAlert, an AI-driven early warning system that integrates fine-tuned BERT-based natural language processing models with unsupervised anomaly detection framework to identify unusual wildlife health events using real-time pre-diagnostic clinical data from wildlife rehabilitation organizations. The NLP module achieved high accuracy across clinical classifications and circumstances of admissions, enabling a pre-diagnostic syndromic surveillance framework. Retrospective validation demonstrated that WildAlert anomalies frequently coincided with or preceded confirmed morbidity events, including highly pathogenic avian influenza (HPAI), harmful algal bloom-associated toxicosis, cold-stunning in sea turtles, mass stranding events, West Nile virus, and mycoplasmosis. WildAlert establishes the worlds largest standardized, near real-time wildlife health surveillance system, transforming wildlife rehabilitation clinical records into actionable intelligence capable of detecting anomalies across taxa and regions, often before other surveillance methods. WildAlert provides a transferable analytical framework and scalable One Health model linking biodiversity monitoring, zoonotic disease preparedness, and ecosystem-linked environmental threats, with implications for conservation, public health and environmental hazard response.

ecology↗

bioRxiv2025

Foundational Architecture Enabling Federated Learning for Training Space Biomedical Machine Learning Models between the International Space Station and Earth

The public and commercial space industries are planning longer duration and more distant space missions, including the establishment of a habitable lunar base and crewed missions to Mars. To support Earth-independent scientific and medical operations, such missions can leverage artificial intelligence and machine learning models to assist with crew healthcare, spacecraft maintenance, and other critical tasks. However, transferring large volumes of data between Earth and space for model development consumes valuable bandwidth, is vulnerable to communication disruptions, and may compromise crew safety and data privacy. Federated learning enables model training while keeping data in situ and only transferring model parameters. In this work, we present a flexible, resilient federated learning framework that provides the secure transmission of model updates between Earth and the International Space Station. On March 15, 2024, this framework pioneered the deployment of federated learning in a spaceflight setting, training classifier models between Earth and the ISS using both real biomedical research data and synthetically generated data.

bioinformatics↗

bioRxiv2024

Epigenetic Disordering Drives Stemness, Senescence Escape and Tumor Heterogeneity

Tumor heterogeneity is the substrate for tumor evolution and the linchpin of treatment resistance. Cancer cell heterogeneity is largely attributed to distinct genetic changes within each cell population. However, the widespread epigenome repatterning that characterizes most cancers is also highly heterogenous within tumors and could generate cells with diverse identities and malignant features. We show that high levels of the epigenetic regulator and oncogene, UHRF1, in zebrafish hepatocytes rapidly induced methylome disordering, loss of heterochromatin, and DNA damage, resulting in cell cycle arrest, senescence, and acquisition of stemness. Reducing UHRF1 expression transitions these cells from senescent to proliferation-competent. The expansion of these damaged cells results in hepatocellular carcinomas (HCC) that have immature cancer cells intermingled with fibroblasts, immune and senescent cells expressing high UHRF1 levels, which serve as reservoirs for new cancer cells. This defines a distinct and heterogenous HCC subtype resulting from epigenetic changes, stemness and senescence escape.

cancer biology↗

bioRxiv2024

Comparative analysis of uncoupled succinate production by the FeII/2-oxoglutarate-dependent dioxygenases

Non-heme iron (FeII) and 2-oxoglutarate(2OG)-dependent dioxygenases catalyse a diverse array of biological reactions. These enzymes couple the oxidative decarboxylation of 2OG to the hydroxylation of the substrates. However, in the absence of the substrate, oxidative decarboxylation of 2OG generates succinate. We have determined succinate level by using succinyl-CoA synthetase to monitor this uncoupled decarboxylation of FeII/2OG-dependent dioxygenases and measured the uncoupled 2OG turnover of different FeII/2OG-dependent dioxygenases. We also performed comparative analysis and verified the functionality of human dioxygenase ALKBH6.

biochemistry↗

bioRxiv2023

A multimodal Transformer Network for protein-small molecule interactions enhances drug-target affinity and enzyme-substrate predictions

Most drugs are small molecules, with their activities typically arising from interactions with protein targets. Accurate predictions of these interactions could greatly accelerate pharmaceutical research. Current machine learning models designed for this task have a limited ability to generalize beyond the proteins used for training. This limitation is likely due to a lack of information exchange between the protein and the small molecule during the generation of the required numerical representations. Here, we introduce ProSmith, a machine learning framework that employs a multimodal Transformer Network to simultaneously process protein amino acid sequences and small molecule strings in the same input. This approach facilitates the exchange of all relevant information between the two types of molecules during the computation of their numerical representations, allowing the model to account for their structural and functional interactions. Our final model combines gradient boosting predictions based on the resulting multimodal Transformer Network with independent predictions based on separate deep learning representations of the proteins and small molecules. The resulting predictions outperform all previous models for predicting drug-target interactions, and the model demonstrates unprecedented generalization capabilities to unseen proteins. We further show that the superior performance of ProSmith is not limited to drug-target interaction predictions, but also leads to improvements in other protein-small molecule interaction prediction tasks, the prediction of Michaelis constants KM of enzyme-substrate pairs and the identification of potential substrates for enzymes. The Python code provided can be used to easily implement and improve machine learning predictions of interactions between proteins and arbitrary drug candidates or other small molecules.

bioinformatics↗

bioRxiv2023

Cdk4/6 is activated by DNA hypomethylation to cause DNA rereplication and organ outgrowth defects

Coordinating epigenomic inheritance and cell cycle progression is essential for organogenesis. UHRF1 connects these functions by facilitating maintenance of DNA methylation and cell cycle progression. Here, we provide evidence resolving the paradoxical phenotype of uhrf1 mutant zebrafish embryos that have both activation of pro-proliferative E2F target genes and increased number of hepatocytes in S-phase, but the liver fails to grow. We find that Atr inhibition reduces DNA replication and increases liver size in uhrf1 mutants, suggesting that uhrf1 mutant hepatocytes have replication stress leading to Atr-mediated cell cycle inhibition and dormant origin firing. We uncover persistent Cdk4/6 activation as the mechanism driving uhrf1 mutant hepatocytes into S-phase, activating Atr and restricting hepatic outgrowth. Palbociclib treatment of uhrf1 mutant embryos prevented aberrant S-phase entry, and the DNA damage response. Palbociclib rescued most cellular and developmental phenotypes in uhrf1 mutants, except DNA hypomethylation, transposon activation and the interferon response. Pro-proliferative genes were also activated in a Cdk4/6 dependent fashion in the liver of dnmt1 mutants, suggesting DNA hypomethylation as a mechanism of Cdk4/6 activation. Thus, the developmental defects caused by DNA hypomethylation are attributed to persistent Cdk4/6 activation leading to DNA replication stress, dormant origin firing and cell cycle inhibition, preventing hepatic outgrowth.

developmental biology↗

bioRxiv2022

Effects of Neurofeedback training on performance in laboratory tasks: A systematic review

Neurofeedback procedures are attracting increasing attention in the neuroscience community. Based on the principle that participants, through suitable feedback, may learn to affect specific aspects of their brain activity, neurofeedback has been applied to basic research, translational, and clinical science alike. A large segment of the extant empirical research as well as review articles have focused on the extent to which neurofeedback interventions affect mental health outcomes, cognitive capacity, aging, and other complex behaviors. Another segment has aimed to characterize the extent to which neurofeedback affects the targeted neural processes. At this time, there is no current systematic review of the effects of neurofeedback on healthy participants performance in experimental tasks. Such a review is relevant in this rapidly evolving field because changes in experimental task performance are traditionally considered a hallmark of changing neurocognitive processes, often established in neurotypical individuals. This systematic review addresses this gap in the literature using the PRISMA method, building on earlier reviews on the same topic. Empirical studies using EEG or fMRI to alter brain processes linked to established, well-defined cognitive and affective laboratory tasks were reviewed. Substantial variability was found regarding the nature of the control for placebo effects, the implementation of the feedback, and the neural targets of feedback. Importantly, only a minority of the studies reported statistically meaningful effects of neurofeedback on performance in cognitive and affective tasks. Examining effect sizes and p-values in a subset of studies found no evidence for reporting bias, while also not finding systematic relations between study characteristics such as sample size or experimental control on the one hand and efficacy on the other. Implications for future work are discussed.

neuroscience↗

bioRxiv2022

The substrate scopes of enzymes: a general prediction model based on machine and deep learning

For a comprehensive understanding of metabolism, it is necessary to know all potential substrates for each enzyme encoded in an organisms genome. However, for most proteins annotated as enzymes, it is unknown which primary and/or secondary reactions they catalyze [1], as experimental characterizations are time-consuming and costly. Machine learning predictions could provide an efficient alternative, but are hampered by a lack of information regarding enzyme non-substrates, as available training data comprises mainly positive examples. Here, we present ESP, a general machine learning model for the prediction of enzyme-substrate pairs, with an accuracy of over 90% on independent and diverse test data. This accuracy was achieved by representing enzymes through a modified transformer model [2] with a trained, task-specific token, and by augmenting the positive training data by randomly sampling small molecules and assigning them as non-substrates. ESP can be applied successfully across widely different enzymes and a broad range of metabolites. It outperforms recently published models designed for individual, well-studied enzyme families, which use much more detailed input data [3, 4]. We implemented a user-friendly web server to predict the substrate scope of arbitrary enzymes, which may support not only basic science, but also the development of pharmaceuticals and bioengineering processes.

bioinformatics↗

bioRxiv2020

Rearing medium dictates variability across replicates in untreated and arsenic challenged zebrafish larvae

Reproducibility and consistency are hallmarks of scientific integrity. Biological systems are inherently noisy, posing a challenge to reproducibility. This is particularly relevant to the field of environmental toxicology, where many unaccounted experimental parameters can have a marked influence on the biological response to exposure. Here, we extend the use of zebrafish as a robust toxicological model for studying the effects of inorganic arsenic (iAs) on liver biology. We observed that iAs toxicity in this system is not influenced by important parameters including genetic background, rearing container material or rearing volume but the dose response to iAs is influenced by the rearing medium. We compared mortality as a measure of iAs toxicity to embryos cultured in two standard rearing media: egg water made from dehydrated ocean salts dissolved in water and a defined embryo medium which is a pH adjusted, buffered salt solution. Larvae reared in egg water were more susceptible to iAs compared to those reared in embryo medium. This effect was independent of the pH differences between these solutions. These culture conditions did not cause any difference in the global hepatic transcriptome of control zebrafish. Further, no difference in the expression of genes involved in the unfolded protein response (UPR) in larvae exposed to iAs treatment or in a stress independent system to activate UPR genes by transgenic overexpression of activating transcription factor 6 (nAtf6) in hepatocytes was observed. However, the clutch-to-clutch variation in gene expression was significantly greater in larvae reared in egg water compared to those in embryo medium. These data demonstrate that egg water affects reproducibility across replicates in terms of gene expression and exacerbates iAs mediated toxic response. This highlights the importance of rigorous evaluation of experimental conditions to assure reproducibility.

pharmacology and toxicology↗

bioRxiv2020

Functional near-infrared spectroscopy in conjunction with electroencephalography of cerebellar transcranial direct current stimulation responses in the latent neurovascular coupling space a chronic stroke study

Cerebellar transcranial direct current stimulation (ctDCS) can facilitate motor learning; however, ctDCS effects have not been investigated using portable neuroimaging vis-a-vis lobular electric field strength. This is important since the subject-specific residual architecture for cerebellar interconnections with the cerebral cortex, including the prefrontal cortex (PFC) and the sensorimotor cortex (SMC), can influence the ctDCS effects on the cerebral functional activation. In this study, we investigated functional near-infrared spectroscopy (fNIRS) in conjunction with electroencephalography (EEG) to measure the changes in the brain activation at the PFC and the SMC following virtual reality (VR)-based Balance Training (VBaT), before and after ctDCS treatment in 12 hemiparetic chronic stroke survivors. Furthermore, we performed general linear modeling (GLM) that can putatively associate the lobular electric field strength due to ctDCS priming with the changes in the fNIRS-EEG measures in the chronic stroke survivors. Here, fNIRS-EEG based measures were investigated in their latent space found using canonical correlation analysis (CCA) that is postulated to capture neurovascular coupling. We found that the ctDCS electrode montage, as well as the state (pre-intervention, during intervention, post-intervention), had a significant (p<0.05) effect on the changes in the canonical scores of oxy-hemoglobin (O2Hb) signal measured with fNIRS. Also, skill acquisition during first exposure to VBaT decreased the activation (canonical score of O2Hb) of PFC of the non-lesioned hemisphere in the novices at their first exposure before the ctDCS intervention. Moreover, ctDCS intervention targeting the leg representation in the cerebellum led to a decrease in the canonical scores of O2Hb at the lesioned SMC, which is postulated to be related to the cerebellar brain inhibition. Furthermore, ctDCS electrode montage, as well as the state, had a significant (p<0.05) interaction effect on the canonical scores of log10-transformed EEG bandpower. Our current study showed the feasibility of fNIRS-EEG imaging of the ctDCS responses in the latent neurovascular coupling space that can not only be used for monitoring the dynamical changes in the brain activation associated with ctDCS-facilitated VBaT, but may also be useful in subject-specific current steering for tDCS to target the cerebral fNIRS-EEG sources to reduce inter-individual variability.

neuroscience↗