Biology subjects

Bhaduri, S.

Publications and source records attributed to Bhaduri, S..

7 recordsLinked to original sources

bioRxiv2025

DeepPROTECTNeo: A Deep learning-based Personalized and RV-guided Optimization tool for TCR Epitope interaction using Context-aware Transformers

Background: The development of personalized cancer vaccines relies on accurately identifying neoepitopes capable of eliciting strong immune responses. T cell receptor (TCR)-epitope interactions are fundamental to cancer immunotherapy. Traditional computational approaches focus primarily on epitope-major histocompatibility complex (MHC) binding, often overlooking the critical contribution of TCR binding. Furthermore, the clinical applicability of existing methods is constrained by fragmented pipelines that require separate workflows for variant calling, HLA typing, and independent peptide-MHC (pMHC) or peptide-TCR (pTCR) evaluation stages. Results: We present DeepPROTECTNeo, a unified deep learning framework that integrates genomic variant detection, HLA typing, high-affinity pMHC binding prediction, variant-driven TCR repertoire mining, followed by a hybrid transformer-Convolutional Neural Network dual-branch feature extractor with an explicit cross-attention-based deep learning model for TCR-epitope binding prediction. Our reverse vaccinology-inspired biologically informed architecture integrates Bidirectional Long short-term memory (Bi-LSTM) sequence features, convolutional-attention physicochemical/evolutionary descriptors via gated fusion, and TCR numbered contextual embeddings to enable residue-level interpretable modelling. Under a strict TCR-split strategy, it achieved a mean AUROC of 0.7856 and AUPRC of 0.7932 outperforming six state-of-the-art predictors by 4-5% with tight inter-fold stability. The architecture maintains high robustness against structural hard negatives and imbalanced datasets, successfully recovering 18 of 34 validated high-affinity neoepitopes from a patient-specific cancer cohort. Conclusions: Experiments results demonstrate that DeepPROTECTNeo is a powerful, reliable end-to-end neoantigen prioritization framework that effectively models complex TCR-epitope interfaces directly from clinical sequencing data, providing a robust interpretable foundation to accelerate personalized cancer immunotherapy.

bioinformatics↗

bioRxiv2022

Cantilever-based in vivo measurements in the early Drosophila embryo reveal adiabatic elastic response on developmentally relevant time scales

In order to understand morphogenesis, it is necessary to know the material properties or forces shaping the living tissue. In spite of this need, very few in vivo measurements are currently available. Here, using the early Drosophila embryo as a model, we describe a novel cantilever-based technique which allows for the simultaneous quantification of applied force and tissue displacement in a living embryo. By analyzing data from a series of experiments in which embryonic epithelium is subjected to developmentally relevant perturbations, we conclude that the response to applied force is adiabatic and is dominated by elastic forces and geometric constraints, or system size effects. Crucially, computational modeling of the experimental data indicated that the apical surface of the epithelium must be softer than the basal surface, a result which we confirmed experimentally. Further, we used the combination of experimental data and comprehensive computational model to estimate the elastic modulus of the apical surface and set a lower bound on the elastic modulus of the basal surface. More generally, our investigations revealed important general features that we believe should be more widely addressed when quantitatively modeling tissue mechanics in any system. Specifically, different compartments of the same cell can have very different mechanical properties; when they do, they can contribute differently to different mechanical stimuli and cannot be merely averaged together. Additionally, tissue geometry can play a substantial role in mechanical response, and cannot be neglected.

biophysics↗

bioRxiv2022

An ERAD-independent role for rhomboid pseudoprotease Dfm1 in mediating sphingolipid homeostasis

Nearly one-third of nascent proteins are initially targeted to the endoplasmic reticulum (ER) where they are correctly folded and assembled before being delivered to their final cellular destinations. To prevent the accumulation of misfolded membrane proteins, ER-associated-degradation (ERAD) removes these clients from the ER membrane to the cytosol in a process known as retrotranslocation. Our recent work demonstrates that rhomboid pseudoprotease, Dfm1, is involved in the retrotranslocation of ubiquitinated integral membrane ERAD substrates. To survey for potential interaction partners of Dfm1, we performed protein-proximity labeling by BioID (proximity-dependent biotin identification) followed by mass spectrometry and identified several interacting proteins known to play a role in the sphingolipid biosynthesis pathway. Specifically, we found that Dfm1 physically interacts with the SPOTS complex, which is composed of serine palmitoyltransferase (SPT) enzymes and accessory components and is critical for catalyzing the first rate-limiting step of the sphingolipid biosynthesis pathway. We demonstrate for the first time that Dfm1 has a role in ER export, a function that is independent of Dfm1s canonical ERAD retrotranslocation function. Specifically, we show that loss of Dfm1 results in the accumulation of phosphorylated Orm2 at the ER, suggesting a novel role for Dfm1 in controlling Orm2 export from the ER and its subsequent degradation by EGAD. Moreover, recruitment of Cdc48 by Dfm1, which is critical for its role in ERAD retrotranslocation, is dispensable for Dfm1s role in ER export. Given that the accumulation of human Orm2 homologs, ORMDLs, are associated with many maladies, our study serves as a molecular foothold for understanding how dysregulation of sphingolipid metabolism leads to various diseases.

cell biology↗

bioRxiv2021

Derlin rhomboid pseudoproteases employ substrate engagement and lipid distortion function for retrotranslocation of ER multi-spanning membrane substrates

Nearly one-third of proteins are initially targeted to the endoplasmic reticulum (ER) membrane where they are correctly folded, assembled, and then delivered to their final cellular destinations. In order to prevent the accumulation of misfolded membrane proteins, ER associated degradation (ERAD) moves these clients from the ER membrane to the cytosol; a process known as retrotranslocation. Our recent work in S. cerevisiae has revealed a derlin rhomboid pseudoprotease, Dfm1, is involved in the retrotranslocation of ubiquitinated ERAD membrane substrates. In this study, we sought to understand the mechanism associated with Dfm1s actions and found that Dfm1s conserved rhomboid residues are critical for membrane protein retrotranslocation. Specifically, we identified several retrotranslocation-deficient Loop 1 mutants that display impaired binding to membrane substrates. Furthermore, Dfm1 has retained the lipid thinning functions of its rhomboid protease predecessors to facilitate in the removal of ER membrane substrates. We find this substrate engagement and lipid thinning feature is conserved in its human homolog, Derlin-1. Utilizing interaction studies and molecular dynamics simulations, this work reveals that derlin rhomboid pseudoproteases employ novel mechanisms of substrate engagement and lipid thinning for catalyzing extraction of multi-spanning membrane substrates.

cell biology↗

bioRxiv2020

Comprehensive analysis of fungal G1 cyclin docking motif sequences that control CDK regulatory potency in vivo

Cyclin-dependent kinases (CDKs) control the ordered series of events during eukaryotic cell division. The stage at which individual CDK substrates are phosphorylated can be dictated by cyclin-specific docking motifs. In budding yeast, substrates with Leu/Pro-rich (LP) docking motifs are recognized by Cln1/2 cyclins in late G1 phase, yet the key sequence features of these motifs and the conservation of this mechanism were unknown. Here we comprehensively analyzed LP motif requirements in vivo by combining a competitive growth assay with mutational scanning and deep sequencing. We quantified the impact of all single-residue replacements in five different LP motifs, using six distinct G1 cyclins from diverse fungi including medical and agricultural pathogens. The results reveal the basis for variations in potency among wild-type motifs, and allow derivation of a quantitative matrix that predicts the potency of other candidate motifs. In one protein, Whi5, we found overlapping LP and phosphorylation motifs with partly redundant effects. In another protein, the CDK inhibitor Sic1, we found that its LP motif is inherently weak due to unfavorable residues at key positions, and this imposes a beneficial delay in its phosphorylation and degradation. The overall results provide a general method for surveying viable docking motif sequences and quantifying their potency in vivo, and they reveal how variations in LP motif potency can tune the strength and timing of CDK regulation.

molecular biology↗

bioRxiv2019

A Novel Chloroplast Super-Complex Consisting of the ATP Synthase and Photosystem I Reaction Center

Several super-complexes of individual hetero-oligomeric membrane protein complexes, whose function is to facilitate intra-membrane electron and proton transfer and harvesting of light energy, have been previously characterized in the mitochondrial cristae and chloroplast thylakoid membranes. The latter membrane is reported here to also be the location of an intra-membrane super-complex which is dominated by the ATP-synthase and photosystem I (PSI) reaction-center complexes, defined by mass spectrometry, clear-native PAGE and Western Blot analyses. This is the first documented presence of ATP synthase in a super-complex with the PSI reaction-center located in the non-appressed stromal domain of the thylakoid membrane.

biochemistry↗

bioRxiv2017

Sporosarcina pasteurii can form nanoscale crystals on cell surface

Using a semi-solid 0.5% agar column, we study the phenomenon of microbially induced mineral (calcium carbonate) by the bacteria Sporosarcina pasteurii. Our platform allows for in-situ visualization of the phenomena, and we found clear evidence of bacterial cell surface facilitating formation of nanoscale crystals. Moreover, in the bulk agar we found the presence of microspheres, which seem to arise from an aggregation of nanoscale crystals with needle like morphology. Extensive chemical characterization confirmed that the crystals to be calcium carbonate, and two different polymorphs (calcite and vaterite) were identified.

biophysics↗