Biology subjects

Bagler, G.

Publications and source records attributed to Bagler, G..

8 recordsLinked to original sources

bioRxiv2025

AllerStack: Predicting Allergenic Proteins with a Stacked Ensemble Approach

Accurate prediction of protein allergenicity is essential for ensuring food and drug safety. While machine learning and deep learning models have been explored for this task, limitations remain in dataset scale, feature representation, and model architecture. Here, we introduce AllerStack, a two-stage stacked ensemble model integrating handcrafted and ESM2-based learned features for allergenicity classification. The model was developed using a balanced dataset comprising 11,930 allergenic and 11,930 non-allergenic proteins. We extracted amino acid composition, dipeptide composition, and physicochemical features using the Biopython library, alongside contextual embeddings from the pre-trained ESM2 protein language model. Diverse classifiers (QDA, SVM, KNN, and ANN) were trained separately on these features in the base layer. Their predictions were used as input to a meta-classifier based on XGBoost. AllerStack achieved high predictive performance with 96.87% accuracy, 96.86% F1-score, 93.75% Matthews correlation coefficient (MCC), and an AUC of 0.99. A publicly accessible web server (https://cosylab.iiitd.edu.in/allerstack/) enables real-time allergenicity prediction from protein sequences. AllerStack provides a robust, interpretable, and user-friendly platform for allergen detection in computational biology. HighlightsO_LIAn extensive dataset of 23,860 proteins. C_LIO_LICombines handcrafted and ESM2-derived features. C_LIO_LIStacked ensemble model with an accuracy of 96.87%. C_LIO_LISHAP-based model interpretability at the model and feature level. C_LIO_LIPublic web server, AllerStack (https://cosylab.iiitd.edu.in/allerstack/). C_LI

bioinformatics↗

bioRxiv2021

Graph Signal Processing on protein residue networks helps in studying its biophysical properties

Understanding the physical and chemical properties of proteins is vital, and many efforts have been made to study the emergent properties of the macro-molecules as a combination of long chains of amino acids. Here, we present a graph signal processing based approach to model the biophysical property of proteins. For each protein inter-residue proximity-based network is used as basis graph and the respective amino acid properties are used as node-signals. Signals on nodes are decomposed on networks Laplacian eigenbasis using graph Fourier transformations. We found that the intensity in low-frequency components of graph signals of residue features could be used to model few biophysical properties of proteins. Specifically, using our approach, we could model protein folding-rate, globularity and fraction of alpha-helices and beta-sheets. Our approach also allows amalgamation of different types of chemical and graph theoretic properties of residue to be used together in a multi-variable regression model to predict biophysical properties.

biophysics↗

bioRxiv2020

A novel cross talk of AtRAV1, an ethylene responsive transcription factor with MAP kinases imparts broad spectrum disease resistance in plants

Plant diseases pose a serious threat to sustainable agriculture as controlling them in eco-friendly manner remains a challenge. In this study, we establish RAV1 as a master transcriptional regulator of defense genes in model plant Arabidopsis. The overexpression of AtRAV1 provided disease resistance against necrotrophic fungal pathogen (Rhizoctonia solani) infection in A. thaliana. The transgenic lines exhibited enhanced expression of several defense genes including mitogen associated protein kinases (MAPKs) and the amplitude of their expression was further enhanced upon pathogen infection. Conversely, the atrav1 mutant plants were unable to induce the expression of these defense genes and were highly susceptible to infection. Our data suggests that upon pathogen attack, AtRAV1 transcriptionally upregulate the expression of MAPKs (AtMPK3, AtMPK4 and AtMPK6) and AtMPK3 and AtMPK6 are essential for AtRAV1 mediated disease resistance. Further, we demonstrate that AtRAV1 is a phosphorylation target of AtMPK3 (but not AtMPK6) and the phospho-defective variants of AtRAV1 are unable to induce disease resistance in A. thaliana. Considering the presence of AtRAV1 orthologs in diverse plant species, we propose that they can be gainfully deployed to control economically important diseases. In deed we observe that overexpression of tomato ortholog of AtRAV1 (SlRAV1) provides broad spectrum disease resistance against bacterial (Ralstonia solanacearum), fungal (R. solani) and viral (Tomato leaf curl virus) infections in tomato.

plant biology↗

bioRxiv2019

Supervised learning of protein thermal stability using sequence mining and distribution statistics of network centrality

MotivationIt is expected that the difference in the thermal stability of mesophilic and thermophilic proteins arises, in part at least, from the differences in their molecular structures and amino acid compositions. Existing machine learning approaches for supervised classification of proteins rely on the features derived from the structural networks and the amino acid sequences. However, the network features used leave out several important network centrality values, the statistic used is a simple average and the sequence features used are hand-picked leading to an accuracy of 90%.\n\nResultsWe show that discriminating sub-sequences of the amino acid sequences can significantly improve classification accuracy compared to the existing approaches of counting amino acids, di-peptide or even tri-peptide bonds. We identify notions of network centrality, specifically that depends on the distances between C atoms, that appears to correlate better with thermal stability compared to the existing network features. We also show how to generate better statistics from the node- and edge-wise centrality values that more accurately captures the variations in their values for different types of proteins. These improved feature selection techniques make it possible to classify between thermophilic and mesophilic proteins with 96% accuracy and 99% area under ROC.\n\nAvailabilityThe dataset and source code used are available at https://github.com/ankits0207/Protein_Classification_BIO699\n\nContactdbera@iiitd.ac.in\n\nonline.

bioinformatics↗

bioRxiv2018

BitterSweet: Building machine learning models for predicting the bitter and sweet taste of small molecules

The dichotomy of sweet and bitter tastes is a salient evolutionary feature of human gustatory system with an innate attraction to sweet taste and aversion to bitterness. A better understanding of molecular correlates of bitter-sweet taste gradient is crucial for identification of natural as well as synthetic compounds of desirable taste on this axis. While previous studies have advanced our understanding of the molecular basis of bitter-sweet taste and contributed models for their identification, there is ample scope to enhance these models by meticulous compilation of bitter-sweet molecules and utilization of a wide spectrum of molecular descriptors. Towards these goals, based on structured data compilation our study provides an integrative framework with state-of-the-art machine learning models for bitter-sweet taste prediction (BitterSweet). We compare different sets of molecular descriptors for their predictive performance and further identify important features as well as feature blocks. The utility of BitterSweet models is demonstrated by taste prediction on large specialized chemical sets such as FlavorDB, FooDB, SuperSweet, Super Natural II, DSSTox, and DrugBank. To facilitate future research in this direction, we make all datasets and BitterSweet models publicly available, and also present an end-to-end software for bitter-sweet taste prediction based on freely available chemical descriptors.

bioinformatics↗

bioRxiv2018

Network controllability: viruses are driver agents in dynamic molecular systems

In recent years control theory has been applied to biological systems with the aim of identifying the minimum set of molecular interactions that can drive the network to a required state. However in an intra-cellular network it is unclear what control means. To address this limitation we use viral infection, specifically HIV-1 and HCV, as a paradigm to model control of an infected cell. Using a large human signalling network comprised of over 6000 human proteins and more than 34000 directed interactions, we compared two dynamic states: normal/uninfected and infected. Our network controllability analysis demonstrates how a virus efficiently brings the dynamic host system into its control by mostly targeting existing critical control nodes, requiring fewer nodes than in the uninfected network. The driver nodes used by the virus are distributed throughout the pathways in specific locations enabling effective control of the cell via the high control centrality of the viral and targeted host nodes. Furthermore, this viral infection of the human system permits discrimination between available network-control models, and demonstrates the minimum-dominating set (MDS) method better accounts for how biological information and signals are transferred than the maximum matching (MM) method as it identified most of the HIV-1 proteins as critical driver nodes and goes beyond identifying receptors as the only critical driver nodes. This is because MDS, unlike MM, accounts for the inherent non-linearity of signalling pathways. Our results demonstrate control-theory gives a more complete and dynamic understanding of the viral hijack mechanism when compared with previous analyses limited to static single-state networks.

systems biology↗

bioRxiv2018

Data-driven analysis of biomedical literature suggests broad-spectrum benevolence of culinary herbs and spices

Spices and herbs are key dietary ingredients used across cultures worldwide. Beyond their use as flavoring and coloring agents, the popularity of these aromatic plant products in culinary preparations has been attributed to their antimicrobial properties. Last few decades have witnessed an exponential growth of biomedical literature investigating the impact of spices and herbs on health, presenting an opportunity to mine for patterns from empirical evidence. Systematic investigation of empirical evidence to enumerate the health consequences of culinary herbs and spices can provide valuable insights into their therapeutic utility. We implemented a text mining protocol to assess the health impact of spices by assimilating, both, their positive and negative effects. We conclude that spices show broad-spectrum benevolence across a range of disease categories in contrast to negative effects that are comparatively narrow-spectrum. We also implement a strategy for disease-specific culinary recommendations of spices based on their therapeutic tradeoff against adverse effects. Further by integrating spice-phytochemical-disease associations, we identify bioactive spice phytochemicals potentially involved in their therapeutic effects. Our study provides a systems perspective on health effects of culinary spices and herbs with applications for dietary recommendations as well as identification of phytochemicals potentially involved in underlying molecular mechanisms.

bioinformatics↗

bioRxiv2018

SpiceRx: an integrated resource for the health impacts of culinary spices and herbs

Spices and herbs are key dietary ingredients used in cuisines across the world. They have been reported to be of medicinal value for a wide variety of diseases through a large body of biomedical investigations. Bioactive phytochemicals in these plant products form the basis of their therapeutic potential as well as adverse effects. A systematic compilation of empirical data involving these aspects of culinary spices and herbs could help unravel molecular mechanisms underlying their effects on health.\n\nSpiceRx provides a platform for exploring the health impact of spices and herbs used in food preparations through a structured database of tripartite relationships with their phytochemicals and disease associations. Starting with an extensive dictionary of culinary spices and herbs, their disease associations were text mined from MEDLINE, the largest database of biomedical abstracts, assisted with manual curation. This information was further combined with spice-phytochemical and phytochemical-disease associations. SpiceRx is an integrated repertoire of evidence-based knowledge pertaining to the health impacts of culinary spices and herbs, and facilitates their disease-specific culinary recommendations as well as exploration of molecular mechanisms underlying their health effects.\n\nAvailability and ImplementationSpiceRx is available at http://cosylab.iiitd.edu.in/spicerx and supports all modern browsers. SpiceRx is implemented with Python web development framework Django and relational database PostgreSQL; the front-end was built using HTML, CSS, JavaScript, AJAX, jQuery, JSME Molecular Editor, Bootstrap, Jmol, DataTables and Google Charts.\n\nSupplementary informationSupplementary data are available at Bioinformatics online.

bioinformatics↗