Biology subjects

Mallik, S.

Publications and source records attributed to Mallik, S..

At least 19 recordsLinked to original sources

bioRxiv2026

A Hybrid Residual-Swin Transformer Design with Attention for Prostate Cancer Segmentation

Prostate cancer, a leading cause of cancer-related deaths among men globally, necessitates the development of precise diagnostic and treatment strategies. Accurate segmentation of prostate cancer in medical imaging, particularly in MRI scans, is crucial for early diagnosis and clinical decisions. Traditional manual segmentation techniques, while efficient, are labor-intensive and necessitate significant expertise, resulting in an increasing demand for automated alternatives. The RSAUNet is a novel deep learning architecture designed to improve prostate cancer segmentation. It incorporates essential components, including Residual Blocks, Swin Transformer Blocks, and Attention Mechanisms within a U-Net architecture. This markedly enhances the model's capacity to discern complex anatomical features and accurately segment malignant tissues. Using sophisticated deep Learning methods, RSAUNet addresses the complexities of prostate imaging, delivering reliable, consistent segmentation results. The model was evaluated against various cutting-edge techniques on extensive multi-parametric MRI datasets, attaining an impressive Dice coefficient (DC) of 0.998 and a Jaccard Index (IoU) of 0.965. These findings highlight the innovative characteristics of RSAUNet and its capacity to transform prostate cancer diagnosis and treatment strategies. The proposed model surpasses current methods and shows potential for practical clinical applications, providing an efficient, precise, and scalable solution for automated prostate cancer segmentation and fostering optimism for the future of medical imaging and diagnosis.

cancer biology↗

bioRxiv2026

Evolution of a genome-architecture-encoded gene regulation system in trypanosomatids

Transcriptional regulation of protein-coding genes is a hallmark of eukaryotic gene expression. Yet, a group of parasitic protists, trypanosomatids, appear to lack this capability. Here, we analyzed genomic, nascent transcriptomic, RNA polymerase occupancy and gene organization data to reconstruct the evolutionary origin and biological consequences of their unusual regulatory strategy. Across 59 Discoba protists, we show stepwise evolutionary erosion of conventional transcription regulation components in trypanosomatida lineage, including gene consolidation into polycistronic transcription units (PTUs), shortening of intra-PTU non-coding regions, and depletion of transcription factors and their enriched DNA-binding motifs. This transition was associated with near-constitutive expression of most genes, indicating broad loss of conditional gene expression. However, trypanosomatids retain some differential regulation at the PTU level, with >70% PTUs featuring significantly different nascent transcription than their neighbors or resident chromosomes. Moreover, gene expression is not uniform within PTUs: nascent transcription, translation efficiency, and protein abundance progressively decline with distance from the transcription start site. Consistent with this architecture-encoded regulatory logic, co-complex subunits and co-pathway enzymes preferentially occupy adjacent positions within PTUs despite each PTUs overall functional heterogeneity. These findings reveal an evolutionary shift from gene-specific transcriptional regulation toward a regime where genome architecture becomes a regulator of gene expression.

evolutionary biology↗

bioRxiv2025

The evolutionary origin of host association and polycistronic transcription in trypanosomatids

Trypanosomatids (Kinetoplastids) encompass multiple lineages of parasitic protists with monoxenous or dixenous life cycles, infecting insects, vertebrates, and plants; in vertebrate hosts, some are intracellular, while others are extracellular. To understand the origin and diversification of their host associations, we integrated comparative genomics across 47 genomes. Results highlight that monoxenous, extracellular trypanosomatids originated from predatory ancestors through reductive evolution, which diminished their metabolic and hunting capabilities. Intracellularity and dixenous lifestyle convergently originated three times independently. Progressive consolidation of genes into polycistronic transcription units (PTUs) was a central innovation that began in early Glycomonada and expanded through chromosome fission-fusion and gene relocation/inversion. In present-day PTUs, protein complex subunits and metabolic pathway enzymes are positioned for co-expression in temporal synchrony, and chromosomes minimize colinear PTUs to counter transcriptional readthrough. Together, these results provide a time-resolved origin of host-association and polycistronic transcription in trypanosomatids, possibly through an intermediate phase of facultative parasitism.

evolutionary biology↗

bioRxiv2025

Short-term sensory memory mediates adaptation, habituation, and a paradoxical neural-behavioral transformation in C. elegans

Repeated exposure to stimuli elicits decreasing sensory neural responses over time (adaptation). However, resulting behavioral responses can either weaken over time (habituation) or remain invariant, indicating that the neural-behavioral link is not fixed. To investigate neural adaptation and its flexible translation into behavioral decision making, we created a mathematical framework hypothesizing (1) that sensory networks optimize the speed and accuracy of encoding exogenous stimuli, and (2) that representations form along two time scales, one embedding immediate information and the other stimulus history. Using experimental recordings of the nematode C. elegans, we validated normative model predictions of this optimal encoding strategy, specifically how neural dynamics and adaptation levels vary with stimulus timing. A parametric Bayesian decoder architecture predicted conditions leading to behavioral habituation or invariance, but also paradoxical inversion, whereby appetitive stimuli elicit aversive responses. Experiments with food odors validated that inversion behavior occurred after several repetitions with a long stimulation time and low odor concentrations. Mechanistically, during sensory neural adaptation, weaker immediate stimulus representations can be compensated by secondary processes through memory effects, with biological origins that remain to be studied.

neuroscience↗

bioRxiv2025

Brain Stroke Prediction: A PSO Optimized Stacked Ensemble Machine Learning Approach with SMOTEEN Data Balancing

Prediction of stroke is a critical challenge in healthcare, where early intervention can significantly reduce risks and improve patient outcomes. Traditional methods often struggle with imbalanced datasets and low prediction accuracy. This paper proposes a novel approach to address these issues by combining PSO-optimized Stacked Ensemble ML Classifiers with SMOTEEN data balancing to predict strokes effectively. The technique optimizes 3 Baseline ML Classifiers(RF, SVM, XGB) using Particle Swarm Optimization (PSO) and Stacked Ensemble approach is applied that significantly improving prediction performance. The proposed methodology achieved an impressive 95.57% accuracy, outperforming conventional models such as SVM, LR, RF, KNN, XGB. Overall, the PSO-optimized stacked ensemble approach outperforms traditional machine learning techniques in stroke prediction, providing a more accurate and robust model for early detection and intervention.

bioengineering↗

bioRxiv2025

Detection and Classification of Cervical Cancer Using Optimized Deep Learning Approach

Given the global impact of the cervical cancer epidemic on peoples health, it is critical to have readily available and effective screening technologies. In order to effectively fight this disease, it is crucial to identify the groups who are most at risk. Our study aims to build a robust deep learning system tailored for the classification of cervical cancer using images acquired from Pap screenings; this will allow us to tackle this challenge head-on. Our approach improves upon previous methods of visual feature detection by using a deep learning model based on transfer learning called Squeeze-and-Excitation-ResNet152. The Deer Hunting Optimization method is used to optimise the network by modifying its hyper-parameters. We test our methods on eleven distinct disease sets, with a total of 8838 images distributed differently. Sources such as CRIC and SIPaKMeD were consulted for the acquisition of these images. In order to reduce dataset bias, we use a cost-sensitive loss function all through training. Impressive performance metrics were generated by the testing set analysis, which significantly outperformed the previous methods. Accuracy was 99.74%, precision was 98.98%, recall was 98.15%, specificity was 98.97%, and F1-Score was 99.06%. We can greatly enhance the identification of problems connected to cervical cancer with our technologies. One way to achieve this goal is to make it easier for medical professionals to make quick diagnosis.

bioengineering↗

bioRxiv2025

Integrative Chemical Genetics Platform Identifies Condensate Modulators Linked to Neurological Disorders

Aberrant biomolecular condensates are implicated in neurological disorders including ALS, frontotemporal dementia, and DYT1 dystonia, yet approaches to systematically identify their modulators remain limited. Here we establish MLF2 as a versatile condensate biomarker and develop CondenScreen, an integrated high-content screening and bioinformatic pipeline enabling identification of condensate modulators across chemical and genetic space. Screening 1,760 FDA-approved compounds in a cellular DYT1 dystonia model, we identify drugs that alter aberrant condensate properties, validating the platform for condensate-targeted drug discovery. In parallel, a genome-wide CRISPR/Cas9 screen links condensate accumulation to microcephaly genes and more than 10 additional neurodevelopmental disorders. Machine learning and confocal imaging resolves distinct condensate phenotypes: loss of microcephaly-associated ZNF335 results in nucleoplasmic condensates, whereas RNF26 deletion produces nuclear envelope condensates that phenocopy hallmarks of torsin deficiency. Our study provides a scalable resource for identifying corrective modulators of condensates and establishes a link between nuclear condensate accumulation and neurodevelopmental disorders.

cell biology↗

bioRxiv2025

Explainable AI-Driven Diagnosis Model for Early Glaucoma Detection Using Grey-Wolf Optimized Extreme Learning Machine Approach

Glaucoma is a prominent threat to vision and ranks as the third leading cause of blindness in India. Early detection is crucial to limit its progression. Retinal image analysis, particularly computer-aided diagnosis (CAD), has gained significant attention due to its potential in effectively screening for and managing glaucoma. Nowadays, Artificial intelligence (AI) has achieved significant progress has been made in medical image analysis across various applications. However, the main challenge hindering the widespread adoption of deep neural models in clinical settings is their lack of interpretability. In response to this need, we propose an enhanced CAD model with four key components: image pre-processing, feature extraction, feature dimensionality reduction, and classification. Utilizing the G1020 and ORIGA datasets, we employ a fast discrete curvelet transform with wrapping (FDCT-WRP) for curve-like feature extraction. A combined feature reduction technique, principal component analysis (PCA) and linear discriminant analysis (LDA), is applied to generate relevant features and reduce the feature dimensionality. Incorporating an improved grey wolf Optimization (IMGWO) with an Extreme Learning Machine (ELM) for classification. Then, IMGWO optimizes parameters, enhancing the efficiency of single-hidden-layer feedforward neural networks. Through a 10 x 5-fold stratified cross-validation (SCV) test on standard datasets, our CAD model demonstrates superior performance. The proposed CAD model achieved better classification result i.e., 93.87% and 95.38% on the G1020 and ORIGA datasets respectively. Additionally, we have proposed a framework named as GlaucoXAI (Glaucoma explainable artificial intelligence), which utilizes seven advanced explanation methods to improve the interpretability of deep learning models for medical experts trust. GlaucoXAI has been employed on glaucoma detection using fundus images. GlaucoXAIs adaptable design could aid ophthalmologists and medical professionals in detection of glaucoma. The experimental results demonstrate that the proposed model surpasses other existing models in classification accuracy, while significantly reducing the number of features.

bioengineering↗

bioRxiv2025

Identification of the Best Filter-based Feature Selection Techniques for Microarray Datasets

This article basically explores the impact of univariate & multivariate filter-based feature selection methodologies on enhancing the classification performance for the real-life classification problems. Our study considers two univariate filter-based feature selection techniques, namely, Chi-square and Fisher score, as well as two multivariate filter-based feature selection techniques, viz., Symmetrical Uncertainty and Minimum Redundancy-Maximum Relevance (mRMR). These methods are applied to feature selection from Five diverse collections of datasets, including datasets related to Mixed-lineage Leukaemia (MLL), Lung Cancer, Ovarian Cancer, Central Nervous System (CNS), and Colon Cancer. For each feature, fitness values are calculated using the four aforementioned feature selection methods. After that, a stratified 10-fold cross-validation procedure is conducted using Support Vector Machines (SVM) and Multilayer Perceptrons (MLP) to determine the classification accuracy for each feature. A set of five microarray datasets was used in this evaluation in order to assess the effectiveness of the filter methods. The results of this study represent the first comprehensive analysis and comparison of gene expression datasets filtered using a variety of ranking strategies. Among these approaches, entropy-based methods (e.g., mRMR) emerge as the most effective. The mRMR method demonstrates Outstanding performance outcomes of accuracy, F1-score, and Root Mean Square Error (RMSE). When comparing classifier performance, the F1-score, which combines precision and recall, is particularly useful, while the RMSE measures prediction accuracy. Chi-square, Fisher Score, and Symmetrical Uncertainty (SU) follow as the second, third, and fourth best approaches, respectively. Although the SVM classifier demonstrates superior performance, the difference in accuracy between SVM and the MLP classifier is marginal. Key PointsO_LIOur study considers two univariate filter-based feature selection techniques, namely, Chi-square and Fisher score, as well as two multivariate filter-based feature selection techniques, viz., Symmetrical Uncertainty and Minimum Redundancy-Maximum Relevance (mRMR). C_LIO_LIThese methods are applied to feature selection from Five diverse collections of datasets, including datasets related to Mixed-lineage Leukaemia (MLL), Lung Cancer, Ovarian Cancer, Central Nervous System (CNS), and Colon Cancer. C_LIO_LIAfter that, a stratified 10-fold cross-validation procedure is conducted using Support Vector Machines (SVM) and Multilayer Perceptrons (MLP) to determine the classification accuracy for each feature. C_LIO_LIAmong these approaches, entropy-based methods (e.g., mRMR) emerge as the most effective. The mRMR method demonstrates Outstanding performance outcomes of accuracy, F1-score, and Root Mean Square Error (RMSE). C_LI

bioinformatics↗

bioRxiv2025

Co-translational assembly promotes functional diversification of paralogous proteins

Homomeric proteins are ubiquitous and mediate myriads of cellular functions. When a gene encoding a homomer duplicates, the resulting paralogs can either form distinct homomers, or evolve into a heteromer containing both paralogs. While such events have extensively shaped proteomes, the molecular mechanisms driving these fates and their associated functional consequences remain largely unknown. Here, we conducted a comprehensive phylogenomic analysis tracing gene duplication histories of 7,377 human paralogs across the eukaryotic lineage and identified their fates using protein interaction data. Simulations and data analyses show that cellular constraints must act as barriers to disfavor heteromerization and promote homomerization. We found that multiple cellular and molecular constraints can serve as barriers, including the lack of co-expression and co-localization. The main barrier, however, is co-translational assembly, which naturally promotes the self-assembly of each paralog from its corresponding mRNA, thus hindering heteromerization. We further established that heteromerization constrains functional divergence, with homomeric paralogs exhibiting stronger signatures of adaptive evolution and functional divergence compared to heteromeric paralogs. Together, these findings identify key biochemical and cellular properties that explain protein function diversification following gene duplication. One Sentence SummaryCo-translational assembly drives the selective homo-oligomerization of paralogs, which in turn promotes their functional divergence. Graphical Abstract O_FIG O_LINKSMALLFIG WIDTH=180 HEIGHT=200 SRC="FIGDIR/small/634331v1_ufig1.gif" ALT="Figure 1"> View larger version (53K): org.highwire.dtl.DTLVardef@b17418org.highwire.dtl.DTLVardef@3d0a61org.highwire.dtl.DTLVardef@1747aeaorg.highwire.dtl.DTLVardef@31b931_HPS_FORMAT_FIGEXP M_FIG C_FIG

evolutionary biology↗

bioRxiv2024

Nodal Modulator (NOMO) is a force-bearing transmembrane protein required for muscle differentiation

The endoplasmic reticulum (ER) relies on specialized membrane-shaping proteins to maintain a continuous network of sheets and tubules that host distinct biological processes. How this intricate structure of the ER membrane system is maintained under conditions of mechanical strain is incompletely understood. NOMO is an ER-resident transmembrane protein that contributes to ER morphology and is highly expressed in striated muscle. In this study, we identify a critical interface between distal Ig domains that enables NOMO to maintain ER morphology and buffer mechanical forces. By incorporating two independent tension sensors in the luminal domain of NOMO, we demonstrate that NOMO assemblies experience forces in the single piconewton (pN) range, with a significant contribution from the identified interface. These newly defined features are important, if not indispensable, for myogenesis, as interface mutations affecting mechanosensing fail to restore the essential role of NOMO during myogenesis in a C2C12 differentiation model. Moreover, NOMO depletion impairs nematode motility, underscoring a broader functional importance in muscle physiology.

cell biology↗

bioRxiv2024

Design and Evaluation of Nanoscale Materials with Programmed Responsivity towards Epigenetic Enzymes

Self-assembled materials capable of modulating their assembly properties in response to specific enzymes play a pivotal role in advancing intelligent encapsulation platforms for biotechnological applications. Here, we introduce a previously unreported class of synthetic nanomaterials that programmatically interact with histone deacetylase (HDAC) as the triggering stimulus for disassembly. These nanomaterials consist of co-polypeptides comprising poly (acetyl L-lysine) and poly(ethylene glycol) blocks. Under neutral pH conditions, they self-assemble into particles. However, their stability is compromised upon exposure to HDACs, depending on enzyme concentration and exposure time. Our investigation, utilizing HDAC8 as the model enzyme, revealed that the primary mechanism behind disassembly involves a decrease in amphiphilicity within the block copolymer due to the deacetylation of lysine residues within the particles hydrophobic domains. To elucidate the response mechanism, we encapsulated a fluorescent dye within these nanoparticles. Upon incubation with HDAC, the nanoparticle structure collapsed, leading to controlled release of the dye over time. Notably, this release was not triggered by denatured HDAC8, other proteolytic enzymes like trypsin, or the co-presence of HDAC8 and its inhibitor. We further demonstrated the biocompatibility and cellular effects of these materials and conducted a comprehensive computational study to unveil the possible interaction mechanism between enzymes and particles. By drawing parallels to the mechanism of naturally occurring histone proteins, this research represents a pioneering step toward developing functional materials capable of harnessing the activity of epigenetic enzymes such as HDACs.

bioengineering↗

bioRxiv2023

Mutational biases contribute to the neutral complexification of protein interaction networks following gene duplication

Biological systems can gain complexity over time. While some of these transitions are likely driven by natural selection, the extent to which they occur without providing an adaptive benefit is unknown. At the molecular level, one example is heteromeric complexes replacing homomeric ones following gene duplication. Here, we build a biophysical model and simulate the evolution of homodimers and heterodimers following gene duplication using distributions of mutational effects inferred from available protein structures. We keep the specific activity of each dimer identical, so their concentrations drift neutrally without new functions. We show that for more than 60% of tested dimer structures, the relative concentration of the heteromer increases over time due to mutational biases that favor the heterodimer. However, allowing mutational effects on synthesis rates and differences in the specific activity of homo- and heterodimers can limit or reverse the observed bias toward heterodimers. Our results show that the accumulation of more complex protein quaternary structures is likely under neutral evolution, and that natural selection would be needed to reverse this tendency.

systems biology↗

bioRxiv2021

Bacterial hopanoids: a newly identified potent inducer of monocyte to macrophage differentiation

Monocyte to macrophage differentiation is an extremely essential cellular and immunological process aimed to combat the assault of foreign pathogens as well as this process has immense importance in the maintenance of physiological homeostasis. Differentiated macrophage, serving as specialized phagocytes is an indispensable component of innate immunity. Besides this, being a well-documented antigen-presenting cell (APC) macrophages also function as a key regulator of both humoral and cell-mediated immunity: the two integral components of adaptive immunity. Bacterial hopanoids are a primitive analogous of eukaryotic sterols that are established as a prominent immunomodulator. In this study, we have demonstrated the isolation protocol of neutral lipid fractions from different Gram-negative bacteria, and adopting several analytical approaches we proposed that the extracted lipid fractions may contain hopanoid as an active component. We have delved deeper into the biological effect of these probable hopanoids like compounds. Here, by considering the several structural and functional attributes like altered phenotype, expression of macrophage-specific markers, increased intracellular organelles, acquirement of enhanced phagocytic and inflammatory property, induction of autophagy, etc we have established hopanoids as a potential inducer of monocyte to macrophage differentiation. Thus, our study has unraveled a new immunomodulator present in Gram-negative bacteria and would undoubtedly help to understand the intricacies of host-pathogen interaction in a better and conclusive manner.

cell biology↗

bioRxiv2020

On the evolution of chaperones and co-chaperones and the exponential expansion of proteome complexity

Across the Tree of Life (ToL), the complexity of proteomes varies widely. Our systematic analysis depicts that from the simplest archaea to mammals, the total number of proteins per proteome expanded ~200-fold. Individual proteins also became larger, and multi-domain proteins expanded ~50-fold. Apart from duplication and divergence of existing proteins, completely new proteins were born. Along the ToL, the number of different folds expanded ~5-fold and fold-combinations ~20-fold. Proteins prone to misfolding and aggregation, such as repeat and beta-rich proteins, proliferated ~600-fold, and accordingly, proteins predicted as aggregation-prone became 6-fold more frequent in mammalian compared to bacterial proteomes. To control the quality of these expanding proteomes, core-chaperones, ranging from HSP20s that prevent aggregation to HSP60, HSP70, HSP90, and HSP100 acting as ATP-fueled unfolding and refolding machines, also evolved. However, these core-chaperones were already available in prokaryotes, and they comprise ~0.3% of all genes from archaea to mammals. This challenge--roughly the same number of core-chaperones supporting a massive expansion of proteomes, was met by (i) higher cellular abundances of the ancient generalist core-chaperones, and (ii) continuous emergence of new substrate-binding and nucleotide-exchange factor co-chaperones that function cooperatively with core-chaperones, as a network.

evolutionary biology↗

bioRxiv2020

Loss of homomeric interactions and heteromers formation is the long-term fate of duplicated homomers

Oligomeric proteins are central to life. Duplication and divergence of their genes is a key evolutionary driver, also because duplications can yield very different outcomes. Given a homomeric ancestor, duplication can yield two paralogs that form two distinct homomeric complexes, or a heteromeric complex comprising both paralogs. Alternatively, one paralog remains a homomer while the other acquires a new partner. However, so far, conflicting trends have been noted with respect to which fate dominates, primarily because different methods and criteria are being used to assign the interaction status of paralogs. Here, we systematically analyzed all Saccharomyces cerevisiae and Escherichia coli oligomeric complexes that include paralogous proteins. We found that the proportions of homo-hetero duplication fates strongly depend on a variety of factors, yet that nonetheless, rigorous filtering gives a consistent picture. In E. coli about 50%, of the paralogous pairs appear to have retained the ancestral homomeric interaction, whereas in S. cerevisiae only [~]10% retained a homomeric state. This difference was also observed when unique complexes were counted instead of paralogous gene pairs. We further show that this difference is accounted for by multiple cases of heteromeric yeast complexes that share common ancestry with homomeric bacterial complexes. Our analysis settles contradicting trends and conflicting previous analyses, and provides a systematic and rigorous pipeline for delineating the fate of duplicated oligomers in any organism for which protein-protein interaction data are available.

evolutionary biology↗

bioRxiv2020

Genome-scale conserved molecular principles of mRNA half-life regulation

Precise control of protein and mRNA degradation is essential for cellular metabolism and homeostasis. Controlled and specific degradation of both molecular species necessitates their engagements with the respective degradation machineries; this engagement involves a disordered/unstructured segment of the substrate traversing the degradation tunnel of the machinery and accessing the catalytic sites. Here, we report that mRNAs comprising longer terminal and/or internal unstructured segments have significantly shorter half-lives; the lengths of the 5' terminal, 3' terminal and internal unstructured segments that affect mRNA half-life are compatible with molecular structures of the 5' exo- 3' exo- and endo-ribonuclease machineries. Sequestration into ribonucleoprotein complexes elongates mRNA half-life, presumably by burying ribonuclease engagement sites under oligomeric interfaces. After gene duplication, differences in terminal unstructured lengths, proportions of internal unstructured segments and oligomerization modes result in significantly altered half-lives of paralogous mRNAs. Side-by-side comparison of molecular principles underlying controlled protein and mRNA degradation unravels their remarkable mechanistic similarities, and suggests how the intrinsic structural features of the two molecular species regulate their half-lives on genome-scale and during evolution.

systems biology↗

bioRxiv2020

Neural Circuit Dynamics for Sensory Detection

We consider the question of how sensory networks enable the detection of sensory stimuli in a combinatorial coding space. We are specifically interested in the olfactory system, wherein recent experimental studies have reported the existence of rich, enigmatic response patterns associated with stimulus onset and offset. This study aims to identify the functional relevance of such response patterns, i.e., what benefits does such neural activity provide in the context of detecting stimuli in a natural environment. We study this problem through the lens of normative, optimization-based modeling. Here, we define the notion of a low dimensional latent representation of stimulus identity, which is generated through action of the sensory network. The objective of our optimization framework is to ensure high fidelity tracking of a nominal representation in this latent space in an energy efficient manner. It turns out that the optimal motifs emerging from this framework possess morphological similarity with prototypical onset and offset responses observed in vivo. Furthermore, this objective can be exactly achieved by a network with reciprocal excitatory-inhibitory competitive dynamics, similar to interactions between principal neurons (PNs) and local neurons (LNs) in the early olfactory system of insects. The derived model also makes several predictions regarding maintenance of robust latent representations in the presence of confounding background information and tradeoffs between the energy of sensory activity and resultant behavioral measures such as speed and accuracy of stimulus detection. Significance StatementA key area of study in olfactory coding involves understanding the transformation from high-dimensional sensory stimulus to low-dimensional decoded representation. Here, we treat not only the dimensionality reduction of this mapping but also its temporal dynamics, with specific focus on stimuli that are temporally continuous. We examine through optimization-based synthesis how sensory networks can track representations without prior assumption of discrete trial structure. We show that such tracking can be achieved by canonical network architectures and dynamics, and that the resulting responses resemble observations from neurons in the insect olfactory system. Thus, our results provide hypotheses regarding the functional role of olfactory circuit activity at both single neuronal and population scales.

systems biology↗