Biology subjects

Paithankar, S.

Publications and source records attributed to Paithankar, S..

11 recordsLinked to original sources

bioRxiv2026

Automating scientific annotations for open transcriptomic profiles via multi-stage agents

Public transcriptomic repositories contain millions of samples, yet their large-scale reuse is hindered by heterogeneous and inconsistently reported metadata. In the Gene Expression Omnibus (GEO), key biological information is often distributed across study- and sample-level records, requiring context-dependent interpretation. Here we present GEOMeta, a large language model (LLM)-based multi-stage workflow with task-specialized agents for automated GEO metadata curation. The pipeline separates metadata retrieval, task-specific information extraction, field standardization, ontology mapping and quality control. Using GEOMeta, we generated standardized annotations for approximately 600,000 human bulk RNA-seq samples. To demonstrate its utility, we benchmarked transcriptome representation models for predicting sex, age, tissue and disease from transcriptome embeddings. We further prospectively annotated newly submitted GEO studies and evaluated 22 frontier LLMs. Recent open-source Flash models achieved annotation quality comparable to leading reasoning models while reducing costs by an order of magnitude. GEOMeta provides a scalable resource and reproducible framework for metadata curation.

bioinformatics↗

bioRxiv2025

Comparative immuno-biology at clinical recognition of early multiple organ dysfunction syndrome in pediatric and adult patients using single-cell transcriptomics

Globally, sepsis remains a major health issue, with Multiple Organ Dysfunction Syndrome (MODS) being a leading cause of mortality. MODS, a severe condition often seen in intensive care units, typically results from infections or trauma and involves complex pathophysiological processes requiring various clinical interventions. Although infections are the main triggers, the mechanisms driving MODS remain unclear. To investigate the transition of sepsis to MODS, we generated a single cell RNA sequencing dataset comprising 86,839 immune cells from pediatric sepsis patients at the clinical onset of MODS patients and age-matched controls, identifying 22 distinct cell types. A cluster of S100 genes, located in the same genomic region, was highly expressed in neutrophils in MODS patients, demonstrating strong diagnostic potential across cohorts (AUC=0.94- 0.99) and potential as therapeutic targets. We found that many B and T cells showed heightened inflammation and increased apoptotic activity during early MODS. Additionally, specific transcription regulators and surface proteins associated with inflammation and S100 regulations were uniquely expressed in MODS. Pseudotime analysis revealed distinct S100 gene expression patterns between controls and MODS. Cell-cell interaction analysis highlighted dendritic cells as key mediators, enhancing communication between plasma cells and V{delta} T cells while activating inflammatory and immunosuppressive pathways. We also analyzed 116,803 immune cells from adult MODS patients, revealing stronger immune dysregulation compared to pediatric MODS, including altered S100 gene expression, and enhanced cell-cell interactions. These findings suggest that S100 genes may serve as a marker for MODS. Furthermore, insights gained from adult MODS could improve our understanding of rare pediatric MODS and contribute to the development of better therapeutics for all MODS patients.

bioinformatics↗

bioRxiv2023

Computational discovery of co-expressed antigens as dual targeting candidates for cancer therapy through bulk, single-cell, and spatial transcriptomics

MotivationBispecific antibodies (bsAbs) that bind to two distinct surface antigens on cancer cells are emerging as an appealing therapeutic strategy in cancer immunotherapy. However, considering the vast number of surface proteins, experimental identification of potential antigen pairs that are selectively expressed in cancer cells and not in normal cells is both costly and time-consuming. Recent studies have utilized large bulk RNA-seq databases to propose bispecific targets for various cancers. But, co-expressed pairs derived from bulk RNA-seq do not necessarily indicate true co-expression of both markers in the same cell. Single-cell RNA-seq (scRNA-seq) can circumvent this issue but the issues in dropouts and low-coverage of transcripts impede the large-scale characterization of co-expressed pairs. ResultsWe present a computational pipeline for bsAbs identification which combines the advantages of bulk and scRNA-seq while minimizing the issues associated with using these approaches separately. We select Hepatocellular Carcinoma (HCC) as a case study to demonstrate the utility of the approach. First, using the bulk RNA-seq samples in the OCTAD database, we identified target pairs that most distinctly differentiate tumors cases from healthy controls. Next, we confirmed our findings on the scRNA-seq database comprising 39,361 healthy cells from vital organs and 18,000 malignant cells from HCC. The top pair was GPC3[~]MUC13, where both genes are co-expressed on the surface of over 30% of malignant HCCs and have very low expression in other cells. Finally, we leveraged the emerging spatial transcriptomic to validate the co-expressed pair in situ. Availability and ImplementationA standalone R package for bsAbs identification in bulk data is available via GitHub (https://github.com/Lionir/bsAbsFinder).

bioinformatics↗

bioRxiv2023

Detection of viral infection in cell lines using ViralCellDetector

Cell lines are commonly used in research to study biology, including gene expression regulation, cancer progression, and drug responses. However, cross-contaminations with bacteria, mycoplasma, and viruses are common issues in cell line experiments. Detection of bacteria and mycoplasma infections in cell lines is relatively easy but identifying viral infections in cell lines is difficult. Currently, there are no established methods or tools available for detecting viral infections in cell lines. To address this challenge, we developed a tool called ViralCellDetector that detects viruses through mapping RNA-seq data to a library of virus genome. Using this tool, we observed that around 10% of experiments with the MCF7 cell line were likely infected with viruses. Furthermore, to facilitate the detection of samples with unknown sources of viral infection, we identified the differentially expressed genes involved in viral infection from two different cell lines and used these genes in a machine learning approach to classify infected samples based on the host response gene expression biomarkers. Our model reclassifies the infected and non-infected samples with an AUC of 0.91 and an accuracy of 0.93. Overall, our mapping- and marker-based approaches can detect viral infections in any cell line simply based on readily accessible RNA-seq data, allowing researchers to avoid the use of unintentionally infected cell lines in their studies.

bioinformatics↗

bioRxiv2023

Bulk RNA-Sequencing of small airway cell cultures from IPF and post-COVID lung fibrosis patients illustrates disease signatures and differential responses to TGF-β1 treatment

IPF is a condition in which an injury to the lung leads to the accumulation of scar tissue. This fibrotic tissue reduces lung compliance and impairs gas exchange. Studies have shown that infection with COVID-19 significantly worsens the clinical outcomes of IPF patients. The exact etiology of IPF is unknown, but recent evidence suggests that the distal small airways, (those having a diameter less than 2 mm in adults), play a role in the early pathogenesis of IPF. TGF-{beta}1 is a main driver of fibrosis in a variety of tissues; the binding of TGF-{beta}1 to its receptor triggers a signaling cascade that results in inflammatory signaling, accumulation of collagen and other components of the extracellular matrix, and immune system activation. This study aimed to investigate possible mechanisms that contribute to worsening lung fibrosis in IPF patients after being diagnosed with COVID-19, with a particular focus on the role of TGF-{beta}1. Small airway cell cultures derived from IPF and post-COVID-19 IPF patient transplant tissues were submitted for RNA-sequencing and differential gene expression analysis. The genetic signatures for each disease state were determined by comparing the differentially expressed genes present in the cells cultured under control conditions to cells cultured with TGF-{beta}1. The genes shared between the culture conditions laid the framework for determining the genetic signatures of each disease. Our data found that genes associated with pulmonary fibrosis appeared to be more highly expressed in the post-COVID fibrosis samples, under both control and TGF-{beta}1-treated conditions. A similar trend was noted for genes involved in the TGF-{beta}1 signaling pathway; the post-COVID fibrosis cell cultures seemed to be more responsive to treatment with TGF-{beta}1. Gene expression analysis, RT-PCR, and immunohistochemistry confirmed increased levels of BMP signaling in the IPF small airway cell cultures. These findings suggest that TGF-{beta}1 signaling in IPF small airway cells could be inhibited by BMP signaling, leading to the differences in genetic signatures between IPF and post-COVID fibrosis.

cell biology↗

bioRxiv2023

Dihydroxy-Metabolites of Dihomo-gamma-linolenic Acid Drive Ferroptosis- Mediated Neurodegeneration

Even after decades of research, the mechanism of neurodegeneration remains understudied, hindering the discovery of effective treatments for neurodegenerative diseases. Recent reports suggest that ferroptosis could be a novel therapeutic target for neurodegenerative diseases. While polyunsaturated fatty acid (PUFA) plays an important role in neurodegeneration and ferroptosis, how PUFAs may trigger these processes remains largely unknown. PUFA metabolites from cytochrome P450 and epoxide hydrolase metabolic pathways may modulate neurodegeneration. Here, we test the hypothesis that specific PUFAs regulate neurodegeneration through the action of their downstream metabolites by affecting ferroptosis. We find that the PUFA, dihomo gamma linolenic acid (DGLA), specifically induces ferroptosis-mediated neurodegeneration in dopaminergic neurons. Using synthetic chemical probes, targeted metabolomics, and genetic mutants, we show that DGLA triggers neurodegeneration upon conversion to dihydroxyeicosadienoic acid through the action of CYP-EH, representing a new class of lipid metabolite that induces neurodegeneration via ferroptosis.

neuroscience↗

bioRxiv2022

Targeting ATP12A, a non-gastric proton pump alpha subunit, for idiopathic pulmonary fibrosis treatment

Idiopathic Pulmonary Fibrosis (IPF) is a pathological condition of unknown etiology which results from injury to the lung and an ensuing fibrotic response that leads to thickening of the alveolar walls and obliteration of the alveolar space. The pathogenesis is not clear and there are currently no effective therapies for IPF. Small airway disease and mucus accumulation are prominent features in IPF lungs, similar to Cystic Fibrosis (CF) lung disease. The ATP12A gene encodes the alpha-subunit of the non-gastric H+, K+-ATPase, which functions to acidify the airway surface fluid and impairs mucociliary transport function in cystic fibrosis patients. We hypothesize that the ATP12A protein may play a role in the pathogenesis of IPF. Our studies demonstrate that ATP12A protein is overexpressed in distal small airways from IPF patient lungs compared to normal human lungs. In addition, overexpression of the ATP12A protein in mouse lungs worsened the bleomycin (BLEO)-induced experimental pulmonary fibrosis. This was prevented by a potassium-competitive proton pump blocker, vonoprazan (VON). This data supports the concept that the ATP12A protein plays an important role in the pathogenesis of lung fibrosis. Inhibition of the ATP12A protein has the potential as a novel therapeutic strategy in IPF.

cell biology↗

bioRxiv2022

TransCell: In silico characterization of genomic landscape and cellular responses from gene expressions through a two-step deep transfer learning

Gene expression profiling of new or modified cell lines becomes routine today; however, obtaining comprehensive molecular characterization and cellular responses for a variety of cell lines, including those derived from underrepresented groups, is not trivial when resources are minimal. Using gene expression to predict other measurements has been actively explored; however, systematic investigation of its predictive power in various measurements has not been well studied. We evaluate commonly used machine learning methods and present TransCell, a two-step deep transfer learning framework that utilizes the knowledge derived from pan-cancer tumor samples to predict molecular features and responses. Among these models, TransCell has the best performance in predicting metabolite, gene effect score (or genetic dependency), and drug sensitivity, and has comparable performance in predicting mutation, copy number variation, and protein expression. Notably, TransCell improved the performance by over 50% in drug sensitivity prediction and achieved a correlation of 0.7 in gene effect score prediction. Furthermore, predicted drug sensitivities revealed potential repurposing candidates for new 100 pediatric cancer cell lines, and predicted gene effect scores reflected BRAF resistance in melanoma cell lines. Together, we investigate the predictive power of gene expression in six molecular measurement types and develop a web portal (http://apps.octad.org/transcell/) that enables the prediction of 352,000 genomic and cellular response features solely from gene expression profiles. Key PointsO_LIWe provide a systematic investigation on evaluating the predictive power of gene expression in six molecular measurement types including protein expression, copy number variation, mutation, metabolite, gene effect score, and drug sensitivity. C_LIO_LITransCell took advantage of the transfer learning technique, showing how to learn knowledge from the source tumors, and transfer learned weight initializations to the downstream tasks in cell lines. C_LIO_LICompared to the baseline methods, TransCell outperformed in metabolite, gene effect score, and drug sensitivity predictions. C_LIO_LITwo cases studies demonstrate that TransCell could identify new repurposing candidates for pediatric cancer cell lines as well as capture the differences of genetic dependencies in melanoma resistant cell lines. C_LI

bioinformatics↗

bioRxiv2021

Published Anti-SARS-CoV-2 In Vitro Hits Share Common Mechanisms of Action that Synergize with Antivirals

The global efforts in the past few months have led to the discovery of around 200 drug repurposing candidates for COVID-19. Although most of them only exhibited moderate anti- SARS-CoV-2 activity, gaining more insights into their mechanisms of action could facilitate a better understanding of infection and the development of therapeutics. Leveraging large-scale drug-induced gene expression profiles, we found 36% of the active compounds regulate genes related to cholesterol homeostasis and microtubule cytoskeleton organization. The expression change upon drug treatment was further experimentally confirmed in human lung primary small airway. Following bioinformatics analysis on COVID-19 patient data revealed that these genes are associated with COVID-19 patient severity. The expression level of these genes also has predicted power on anti-SARS-CoV-2 efficacy in vitro, which led to the discovery of monensin as an inhibitor of SARS-CoV-2 replication in Vero-E6 cells. The final survey of recent drug- combination data indicated that drugs co-targeting cholesterol homeostasis and microtubule cytoskeleton organization processes more likely present a synergistic effect with antivirals. Therefore, potential therapeutics should be centered around combinations of targeting these processes and viral proteins.

bioinformatics↗

bioRxiv2020

Reversal of Infected Host Gene Expression Identifies Repurposed Drug Candidates for COVID-19

The repurposing of existing drugs offers the potential to expedite therapeutic discovery against the current COVID-19 pandemic caused by the SARS-CoV-2 virus. We have developed an integrative approach to predict repurposed drug candidates that can reverse SARS-CoV-2-induced gene expression in host cells, and evaluate their efficacy against SARS-CoV-2 infection in vitro. We found that 13 virus-induced gene expression signatures computed from various viral preclinical models could be reversed by compounds previously identified to be effective against SARS- or MERS-CoV, as well as drug candidates recently reported to be efficacious against SARS-CoV-2. Based on the ability of candidate drugs to reverse these 13 infection signatures, as well as other clinical criteria, we identified 10 novel candidates. The four drugs bortezomib, dactolisib, alvocidib, and methotrexate inhibited SARS-CoV-2 infection-induced cytopathic effect in Vero E6 cells at < 1 {micro}M, but only methotrexate did not exhibit unfavorable cytotoxicity. Although further improvement of cytotoxicity prediction and bench testing is required, our computational approach has the potential to rapidly and rationally identify repurposed drug candidates against SARS-CoV-2. The analysis of signature genes induced by SARS-CoV-2 also revealed interesting time-dependent host response dynamics and critical pathways for therapeutic interventions (e.g. Rho GTPase activation and cytokine signaling suppression).

genomics↗

bioRxiv2020

Ibrutinib blocks YAP1 activation and reverses BRAFi resistance in melanoma cells

Most BRAF-mutant melanoma tumors respond initially to BRAFi/MEKi therapy, although few patients have durable long-term responses to these agents. The goal of this study was to utilize an unbiased computational approach to identify inhibitors which reverse an experimentally derived BRAFi resistance gene expression signature. Using this approach, we found that ibrutinib effectively reverses this signature and we demonstrate experimentally that ibrutinib re-sensitizes a subset of BRAFi-resistant melanoma cells to vemurafenib. Ibrutinib is used clinically as a BTK inhibitor; however, neither BTK deletion nor treatment with acalabrutinib, another BTK inhibitor with reduced off-target activity, re-sensitized cells to vemurafenib. These data suggest that ibrutinib acts through a BTK-independent mechanism in vemurafenib re-sensitization. To better understand this mechanism, we analyzed the transcriptional profile of ibrutinib-treated BRAFi-resistant melanoma cells and found that the transcriptional profile of ibrutinib was highly similar to that of multiple SRC kinase inhibitors. Since ibrutinib, but not acalabrutinib, has significant off-target activity against multiple SRC family kinases, it suggests that ibrutinib may be acting through this mechanism. Furthermore, genes either upregulated or downregulated by ibrutinib treatment are enriched in YAP1 target genes and we showed that ibrutinib, but not acalabrutinib, reduces YAP1 activity in BRAFi-resistant melanoma cells. Taken together, these data suggest that ibrutinib, or other SRC family kinase inhibitors, may be useful for treating some BRAFi/MEKi-refractory melanoma tumors.

cancer biology↗