Biology subjects

Bharanikumar, R.

Publications and source records attributed to Bharanikumar, R..

5 recordsLinked to original sources

bioRxiv2026

DecoderTCR: Compositional Pretraining and Entropy-Guided Decoding for TCR-pMHC Interactions

Modeling recognition between T-cell receptors (TCRs) and peptide-MHC (pMHC) complexes is a fundamental challenge in computational immunology, constrained by sparse paired interaction data relative to abundant unpaired sequences. We introduce DecoderTCR, a masked language model framework that addresses this through two contributions: (1) a compositional continual pre-training curriculum that learns component representations from marginal data before refining cross-chain dependencies from limited pairs, and (2) Iterative Entropy-Guided Refinement (IEGR), a non-autoregressive decoding algorithm that resolves high-confidence positions first to provide context for uncertain regions. On held-out benchmarks, DecoderTCR achieves 0.96 AUROC for zero-shot pMHC binding prediction and 0.76 AUROC for epitope-specific TCR recognition, approaching supervised baselines without epitope-specific training. Learned representations recover structural contacts without coordinate supervision, and generated sequences exhibit realistic recombination statistics. Experimental validation reveals a prediction-generation gap: strong discrimination does not yet yield reliable generation, highlighting an open challenge for the field.

bioinformatics↗

bioRxiv2025

Shared TCRs in peripheral blood offer robust celiac disease classification independent of gluten Intake

Background and AimsCeliac disease (CeD) is a chronic digestive autoimmune disorder affecting approximately 1% of the worldwide population. It is driven by T cells activated by specific HLA-DQ2 or HLA-DQ8 molecules leading to the destruction of intestinal villi. We aimed to characterize shared CeD-specific T cells in patients on and off a gluten-free diet (GFD) from a large cohort of cases and controls. MethodsWe performed bulk TCR{beta} immune sequencing of the peripheral blood of 1,604 biopsy-confirmed CeD patients (1,339 on GFD, 265 on normal diet) and 1,100 controls. We identified over 300 TCR{beta}s enriched in CeD cases versus controls in an HLA-aware manner, controlling for CeD risk alleles. ResultsCeD-associated TCR{beta}s were found to be more predictive of disease than previously characterized gliadin and glutenin-binding TCRs in a validation cohort. Furthermore, the clonal breadth of these TCR{beta}s was associated with increased intestinal damage. Immune sequencing of the peripheral blood also uncovered repertoire-level differences between CeD patients and controls. CeD patients displayed significantly higher productive clonality compared to age-matched controls as well as expansion of TCR{beta}s specific to cytomegalovirus (CMV) and Epstein-Barr Virus (EBV). ConclusionsThese findings underscore the value of unbiased immune repertoire sequencing to identify novel biomarkers for autoimmune disease and to discover new disease mechanisms which can improve both diagnosis and treatment of disease.

immunology↗

bioRxiv2024

Antigen-driven expansion of public clonal T cell populations in inflammatory bowel diseases

BackgroundInflammatory Bowel Diseases (IBDs), including Crohns disease (CD) and ulcerative colitis (UC), are known to involve shifts in the T-cell repertoires of affected individuals. These include a reduction in regulatory T cells in both diseases, increase in TNF production in CD, expansion of an unconventional T-cell population in CD, and clonal expansion of abundant T-cell populations in CD mucosal tissue. There are also differential HLA risk and protective alleles between CD and UC, implying CD- and UC-specific repertoire changes that have not yet been identified. MethodsWe performed ImmunoSequencing on blood samples from 3,853 CD cases, 1,803 UC cases, and 5,596 healthy controls. For each sample we imputed HLA type and cytomegalovirus (CMV) infection status based on public T-cell receptor {beta} (TCRB) usage and identified public TCRBs enriched in CD or UC cases. FindingsWe determine that there is more expansion across clonotypes in CD, but not UC, compared with healthy controls. We also identify novel interactive effects of HLA-DQ heterodimers with CD and UC risk. Strikingly, from blood we identify public TCRBs specifically expanded in CD or UC. These sequences are more abundant in intestinal mucosal samples, form groups of similar CDR3 sequences, and can be associated to specific HLA alleles. Although the prevalence of these sequences is higher in ileal and ileocolonic CD than colonic CD or UC, the TCRB sequences themselves are shared across CD and not between CD and UC. InterpretationThere are peptide antigens that commonly evoke immune reactions in IBD cases and rarely in non-IBD controls. These antigens differ between CD and UC. CD, particularly ileal CD, also seems to involve more substantial changes in clonal population structure than UC, compared to healthy controls.

immunology↗

bioRxiv2019

Riboflow: using deep learning to classify riboswitches with ~99% accuracy

Riboswitches are cis-regulatory genetic elements that use an aptamer to control gene expression. Specificity to cognate ligand and diversity of such ligands have expanded the functional repetoire of riboswitches to mediate mounting apt responses to sudden metabolic demands and signal changes in environmental conditions. Given their critical role in microbial life, and novel uses in synthetic biology, riboswitch characterisation remains a challenging computational problem. Here we have addressed the issue with advanced deep learning frameworks, namely convolutional neural networks (CNN), and bidirectional recurrent neural networks (RNN) with Long Short-Term Memory (LSTM). Using a comprehensive dataset of 32 ligand classes and a stratified train-validate-test approach, we demonstrated the superior performance of both the deep models (CNN and RNN) relative to other conventional machine learning classifiers on all key performance metrics, including the ROC curve analysis. In particular, the bidirectional LSTM RNN emerged as the best-performing learning method for identifying the ligand-specificity of riboswitches with an accuracy > 0.99 and macro-averaged F-score of 0.96. A dynamic update functionality is inbuilt to account for the discovery of new riboswitches and extend the predictive modelling to any number of new additional classes. Our work would be valuable in the design and assembly of genetic circuits and the development of the next generation of antibiotics. The software is freely available as a Python package and standalone resource for wide use in genome annotation and biotechnology workflows. AvailabilityPyPi package: riboflow @ https://pypi.org/project/riboflow Repository with Standalone suite of tools: https://github.com/RiboswitchClassifier Language: Python 3.6 with numpy, keras, and tensorflow libraries. Licence: MIT

bioinformatics↗

bioRxiv2018

PromoterPredict: sequence-based modelling of Escherichia coli σ70 promoter strength yields logarithmic dependence between promoter strength and sequence

We present PromoterPredict, a dynamic multiple regression approach to predict the strength of Escherichia coli promoters binding the {sigma}70 factor of RNA polymerase. {sigma}70 promoters are ubiquitously used in recombinant DNA technology, but characterizing their strength is demanding in terms of both time and money. Using a well-characterized set of promoters, we trained a multivariate linear regression model and found that the log of the promoter strength is significantly linearly associated with a weighted sum of the -10 and -35 sequence profile scores. It was found that the two regions contributed almost equally to the promoter strength. PromoterPredict accepts -10 and -35 hexamer sequences and returns the predicted promoter strength. It is capable of dynamic learning from user-supplied data to refine the model construction and yield more confident estimates of promoter strength.\n\nAvailabilityOpen source code and a standalone executable with both dynamic model-building and prediction are available (under GNU General Public License 3.0) at https://github.com/PromoterPredict, and require Python 2.7 or greater. PromoterPredict is also available as a web service at https://promoterpredict.com.\n\nContactapalania@scbt.sastra.edu

bioinformatics↗