Mathematical Biosciences and Engineering

This paper presents a unified six-state Continuous-Time Markov Chain (CTMC) framework for Chronic Kidney Disease (CKD) progression, with CKD stages 1-5 modeled as transient states and death as an absorbing state. Under a non-homogeneous CTMC formulation, we derive integral representations for transition probabilities, state distributions, sojourn times, and survival-related quantities. We then study the homogeneous case as a tractable baseline and provide explicit formulas for key quantities. Although the methodology is rooted in standard multi-state theory, these expressions are often left implicit in applied analyses; here they are written out explicitly within a unified CKD framework. We construct covariate-dependent transition rates through a proportional hazards structure, using the standard identification of cause-specific hazards with CTMC transition rates. We fit the time-homogeneous baseline model to 335,283 longitudinal observations from 21,100 synthetic electronic health record patients by maximum likelihood. In this synthetic cohort, the covariate model improves held-out log-likelihood per transition over the null model, with stable performance across 10-times-repeated 5-fold cross-validation, and reproduces the main population-level prevalence patterns. The transition-specific estimates can also be translated into sojourn-time and survival summaries. The model suggests that male sex is associated with faster progression across nearly all CKD transitions, and that hypertension shows a stage-dependent association, with lower estimated transition rates in early stages but a substantial acceleration of the Stage 4 to Stage 5 transition. Overall, the proposed framework provides a mathematically explicit approach for studying CKD trajectories from longitudinal health records.

Ordinary differential equation models of biochemical reactions are often formulated as stoichiometric systems in which the dynamics arise from a collection of interacting processes. A central challenge is that the functional form of each process is rarely known a priori and may be difficult to infer from data. We propose biochemically informed neural ordinary differential equations (BINODEs), a neural-ODE framework that retains the stoichiometric structure of mechanistic models while representing individual processes by neural networks. In BINODEs, the outputs of neural network processes (NNPs) are mapped to state derivatives through a linear layer analogous to a stoichiometric matrix. This architecture allows biological side information, such as process-specific inputs, sign constraints, and monotonicity assumptions, to be built directly into the model. We characterize the approximation properties of NNPs for several standard biochemical rate laws and show that the proposed framework recovers both trajectories and process-level structure in Monod, Lotka-Volterra, pharmacokinetic, and ultradian endocrine models. These results suggest that BINODEs offer a useful compromise between mechanistic interpretability and data-driven flexibility for modeling partially known biochemical or biological dynamical systems.

The Hajj is an annual pilgrimage made by millions of Muslims to Mecca in the Kingdom of Saudi Arabia (KSA). The large number of international attendees at the Hajj increases the risk of global infectious disease spread. However, we know very little about the benefits, costs, and cost-effectiveness of testing and quarantining strategies to contain epidemic spread during mass gathering events. In this work we developed a stochastic discrete-time compartmental metapopulation model to simulate international epidemics of infectious pathogens and their potential importation into KSA during the Hajj. We used the model and an epidemic simulation study to evaluate the impact and cost-effectiveness of three testing and quarantining strategies for arriving pilgrims: randomly testing 99% of pilgrims, 80% of pilgrims, or using a symptom-based screening strategy. The simulations lasted 100 days, covering the 30 days before the Hajj and 65 days after the Hajj. Under the conditions assumed in our simulation study, there was strong evidence that testing and quarantining strategies are cost-effective measures for controlling epidemic threats at the Hajj. The median net monetary benefits of intervention strategies ranged from Intl$-41.89M [95% quantile range Intl$-42.37M to Intl$3.18B] to Intl$12.68B [Intl$-8.70B to Intl$13.82B] across scenarios with different pathogen characteristics (based on the natural histories of SARS-CoV-2 and H1N1 Influenza) and epidemic seed locations. Our results were sensitive to the data sources that were used to estimate the number of pilgrims travelling to KSA by origin country, with flight passenger statistics providing biased estimates of pilgrim numbers. Our work provides an adaptable tool to inform infectious disease risk assessments and evaluate the cost-effectiveness of possible disease control measures for the Hajj, and could be extended to other mass gathering events.

Physical activity (PA) plays an important role in maintaining and improving health. Daily steps have been a key PA measure that is easily accessible with common wearable devices. However, methods are lacking to recommend a personalized optimal distribution of daily steps over a period of time for the best of certain health biomarkers. In this paper, we fill this void based on the data from the All of Us Research Program which includes months of step counts as well as repeated measurements of key health biomarkers. We develop a new offline reinforcement learning (RL) algorithm to learn personalized and optimal PA distributions associated with cardiometabolic risk, where the action is a function representing the daily step distribution over a period of time. Simulation studies demonstrate the advantage of the proposed approach over existing continuous-action RL methods. The learned optimal policy from the All of Us data generally suggests people take more daily steps and also follow a more consistent pattern of PA over time while offering tailored recommendations for subgroups in blood glucose level, body mass index, blood pressure, age, and sex.

Embryo implantation is early and complex stage of pregnancy begins when competent blastocyst makes a physiological attachment to receptive endometrium. Expression of numerous molecules are essential for initiation of pregnancy. leukemia inhibitory factor (LIF) is essential cytokines required for priming uterus to make it receptive for implantation. In mice, the ovarian estrogen regulated expression of LIF is absolutely required for implantation. Golden hamster showed ovarian estrogen independent process of embryo implantation. Hence, the regulation of LIF in uterus of golden hamster during early pregnancy is still ambiguous. In this study, we explored the possible regulation of LIF by uterine factor and their spatio-temporal localization and expression in the uterus of golden hamster during early pregnancy and pseudopregnancy. We further demonstrated their ability to activate prostaglandin synthesizing enzymes to achieve successful pregnancy. We used immunohistochemistry, quantitative and semiquantitative PCR to achieve the objectives. We observed the expression of LIF in all the day of early pregnancy and pseudopregnancy in the uterus of hamster. Their m-RNA was found to be upregulated around the day of implantation and decidualization. LIF showed high expression in D3 pseudopregnancy. LIF was found to be regulated by estrogen in ovariectomized uterus and significantly reduced expression of LIF was observed in letrozole treated uterine horn. Downregulated expression of prostaglandin synthesizing enzymes was observed in anti-LIF antibody treated uterus. Together, these findings highlights that uterine factor regulated LIF mediate their action via activating prostaglandin synthesizing enzymes to make uterus receptive for successful early pregnancy in hamster. HighlightO_LIExpression of LIF in uterus during pregnancy in golden hamster is independent from the presence of blastocyst C_LIO_LILIF is regulated by estrogen in ovariectomized hamster C_LIO_LIExpression of LIF mRNA is downregulated in letrozole treated uterine horn in day 5 of pregnancy indicating the possibility of their regulation by uterine estrogen in golden hamster C_LIO_LIProstaglandin synthesizing enzyme and LIF might be associated with the activation of inflammatory signals which are essential for successful establishment of early pregnancy in golden hamster. C_LI

This work focuses on the global and partial identification problem for fractional differential equations. We provide a general numerical procedure based on global and local optimization algorithms with two refinements for biological systems that ensure solution positivity and homogeneous parameter units. The method is applied to a new fractional model of Dengue outbreak called the Fractional Homogeneous Nishiura (FHN) model, calibrated using data of newly infected people in Cape Verde. We show that our identification method yields a better fit between data and model solutions than previous approaches and that our FHN model captures the dynamics of Dengue more closely than existing systems.

Ectotherm insects growth and development are dictated by temperature and humidity. Conducive habitats and the availability of resources set ideal conditions for insect population growth. Mosquitoes require water, favorable temperature, and blood meal to survive. In this research, we picked a rapidly growing megacity, Ahmedabad, in western India, to explore and establish potential linkages between disease spread and meteorological conditions. Ahmedabad, with a population of over 8 million, is experiencing changes in rain and humidity patterns, pushing the city towards changing vector-borne disease dynamics. We examined dengue cases over ten years, 2012-22, and explored their connections with two prominent climatic variables, temperature and relative humidity. Our findings indicate that stable temperature (25-27.5 {degrees}C) and humidity (> 60%) interaction is a ruling factor in spikes in dengue cases in the city. While stable temperature ranges triggers the dengue cases, RH drives the explosive phases and sustainability of such episodes. Statistically significant increasing trends in temperatures, narrowing down of the day-night temperature ranges, and increasing night temperatures provide more stable temperature regimes in a warming world thereby likely to extend the dengue season beyond the usual monsoon season. Plain Language SummaryDengue incidences have been found to be associated with mosquito population outbreaks. Every year, thousands of lives are lost due to this deadly virus spread by mosquitoes. Particularly in the Indian subcontinent, a large proportion of these cases is associated with the monsoon season and rain patterns. In recent years, there have been abrupt spikes in dengue cases across Indian cities, particularly in western India. To understand this complex interaction of viral proliferation and local environmental conditions, the last ten years of dengue case patterns have been scanned in parallel to the climate data. Our findings suggest that stable temperature windows and humidity levels above certain thresholds trigger a rise in dengue cases. While stable temperature ranges trigger dengue cases, humidity drives such episodes explosive phases and sustainability. Our work pinpoints specific temperature-humidity combinations and suggests that local municipal corporations use them as warning indicators to initiate preventive measures.

Biological pathway analysis often requires identifying interventions that block reachability to an undesirable state, such as a disease-associated module, toxic byproduct, or adverse phenotype, while preserving reachability among essential biological functions. Motivated by this setting, we study the Reachability Preserving Minimum Edge Cut (RPMEC) problem: given protected terminals s1 and s2 and a target terminal t, the goal is to remove a minimum-cost set of edges that separates s1 and s2 from t while keeping s1 and s2 connected. This formulation naturally models pathway-level intervention design, where one seeks to disrupt harmful signaling, metabolic, or interaction routes without breaking required functional connectivity. We revisit the three-terminal undirected edge-cut case and analyze a Dijkstra-style dynamic programming algorithm that is exact on planar graphs but fails on general graphs. We characterize the structural condition required for exactness, namely frontier-realizability of optimal source-side regions, and identify biological graph representations where this condition is likely to hold after appropriate preprocessing, including curated planar pathway maps, Reactome-style hierarchy trees, SCC-contracted feedback modules, metabolic building-block DAGs with dominator structure, and functional-module quotients of protein interaction networks. We further discuss directed variants, approximation strategies, and exact alternatives based on ASP, MILP, bounded-treewidth dynamic programming, and important separators. The results provide a graph-theoretic foundation for deciding when fast greedy computation is reliable for biological pathway intervention problems and when more expressive exact optimization methods are needed. Author SummaryMany real-world networks require interventions that separate harmful or undesirable states while preserving essential connectivity. This situation appears in biological pathway analysis, where one may want to block reachability to a disease-related module, toxic byproduct, or adverse phenotype without disrupting communication among essential genes, proteins, reactions, or metabolites. We study this problem through the Reachability Preserving Minimum Edge Cut formulation. Unlike ordinary minimum cut, the solution must satisfy both a separation requirement and a preservation requirement. We show why a natural Dijkstra-style algorithm works only under specific structural conditions, such as planar, laminar, or module-like pathway graphs, and why it may fail on general graphs. The results help identify when fast graph algorithms are reliable for biological intervention problems and when exact optimization tools such as Answer Set Programming or integer programming are more appropriate.

Plant disease detection using deep learning is essential for precision agriculture, enabling early and automated crop health monitoring. This study proposes an end-to-end transfer learning pipeline, LeafyVGG-16, for multi-class classification of plant diseases and nutrient deficiencies using a tomato leaf dataset. The framework integrates data preprocessing, augmentation, and a VGG-16 backbone with a two-stage fine-tuning strategy. The proposed model is evaluated against CNN, DenseNet-121, Inception-V3, EfficientNetB0, and ResNet-50, achieving an accuracy of 0.93 with precision, recall, and F1-scores of 0.93, 0.90, and 0.92, respectively. These results demonstrate the effectiveness of transfer learning for fine-grained plant disease recognition. We further evaluate model robustness under adversarial cyber attacks to assess deployment reliability in agricultural systems. Under Fast Gradient Sign Method (FGSM) attacks ({epsilon} = 0.01- 0.05), the model shows an accuracy drop of 1%-7.5%, while Projected Gradient Descent (PGD) attacks ({epsilon} = 0.05, step size = 0.005, 10 iterations) produce similar degradation, highlighting the models vulnerability to adversarial perturbations. These findings highlight potential security and reliability risks in AI-based agricultural decision-making systems. Future work will focus on improving robustness and cyber-resilience and extending this framework to other crops for secure and context-aware deployment in resource-constrained environments.

An ongoing outbreak of Bundibugyo virus disease (BVD) in the Democratic Republic of the Congo was deemed a public health emergency of international concern in May 2026. To prevent cross-border importation, many countries, including the United States, Canada, India, Thailand, and Kenya have already proposed containment strategies, and others are likely to follow suit. How well (or poorly) are screening and quarantine containment measures are likely to work? We leverage established epidemiological theory and develop a mathematical model of traveler screening and post-arrival quarantine for BVD to answer this question. We find that traveler screening via symptom screening or molecular testing will miss the majority of infected travelers, and should be complemented by post-arrival quarantine and monitoring of sufficient duration to detect those with long incubation periods. Our findings underscore the limitations of border screening and the importance of complementary measures like post-arrival quarantine to prevent local importation of BVD.

Recent studies have shown that implementing explicit social cues, such as gaze, facial expressions, and gestures, in artificial agents can improve impressions of these agents. However, humans may also use implicit physiological cues, such as facial coloration and cardiac information, in social perception. The present study examined whether subtle skin color changes reflecting pulse signals enhance the perceived human likeness of artificial agents, and whether this effect depends on agent type, signal type, observers interoceptive sensibility, and their awareness of the skin color changes. Participants observed morphed face stimuli created from artificial agents and human faces and judged whether each stimulus appeared human-like or robot-like. In Experiment 1, skin color changes based on human-derived pulse wave signals enhanced perceived human likeness for a highly human-like agent, but not for a less human-like agent. In Experiment 2, perceived human likeness was enhanced not only by pulse-based skin color changes but also by sinusoidal skin color changes matched to the pulse wave signal in terms of mean amplitude and number of peaks. In addition, participants with higher scores on some subscales of the Multidimensional Assessment of Interoceptive Awareness (MAIA), a subjective measure of interoceptive sensibility, tended to notice the skin color changes. However, neither observers interoceptive sensibility nor their awareness of skin color changes directly explained the enhancement of perceived human likeness induced by skin color changes. These results suggest that subtle skin color changes reflecting pulse wave information may function as implicit dynamic cues signaling embodiment or biologicalness in artificial agents, thereby contributing to perceived human likeness.

Malaria transmission in Nigeria is highly seasonal and climate-sensitive, yet routine surveillance and meteorological datasets remain underutilized for predictive modelling at subnational levels. This study modelled seasonal malaria incidence trends in Nasarawa State, Nigeria using routine surveillance and climatic data. A retrospective ecological time-series study was conducted using monthly confirmed malaria incidence data from all 13 Local Government Areas of Nasarawa State between 2021 and 2025. Rainfall and temperature were examined as the climatic predictors. Seasonal decomposition and cross-correlation analyses were performed to identify the temporal patterns and lag structures. Seasonal Autoregressive Integrated Moving Average (SARIMA) and Seasonal Autoregressive Integrated Moving Average with Exogenous Variables (SARIMAX) models were developed using the Box-Jenkins framework. Model performance was evaluated using the Root Mean Square Error (RMSE) and Mean Absolute Percentage Error (MAPE). Malaria incidence showed pronounced seasonal peaks, with the highest transmission occurring during the rainy season. Cross-correlation analysis identified rainfall at a one-month lag and contemporaneous temperature as significant predictors of malaria incidence. The SARIMAX model outperformed the univariate SARIMA model, achieving strong predictive accuracy (MAPE = 8.7%). Forecast projections indicate sustained transmission with a peak incidence expected between June and August 2026. Malaria transmission in Nasarawa follows a predictable seasonal pattern that is influenced by climatic variability. Incorporating rainfall and temperature into SARIMAX models improves the forecasting performance and provides evidence supporting climate-informed malaria surveillance and preparedness in endemic settings.

Male infertility has emerged as a significant concern in modern society, with genetic defects as one of the major underlying cause behind it. This impairment negatively impacts sperm motility and morphology, leading to conditions such as Asthenozoospermia (reduced sperm motility), Teratozoospermia (abnormal sperm morphology) and sometimes Asthenoteratozoospermia (both motility and morphology defects). Assisted reproductive technologies (ART), such as in-vitro fertilization (IVF), offer a potential solution for such cases but with a low success rate. Classical semen analysis provides only a phenotypic snapshot without revealing the fertilizing potential of the sperms. Hence, in order to screen the functional sperm population as well as to get a deeper insight into the reasons underlying the aberrant sperm population, it is important to study their genetic profile. In this work, we have performed a meta analysis of the transcriptomic data of infertile sperms from Asthenozoospermia and Teratozoospermia patients with that from fertile sperms of normal individuals. Thereafter we have screened a signature gene set which has been used to develop a prediction model named Explainable Infertility Test (E-InfertilityTest) to classify between fertile versus infertile sperm at the preliminary level. For each prediction, it will also provide the set of genes which are playing a dominant role towards such prediction. Thus, it will provide patient specific dominant gene expression profile responsible for the aberration. This work warrants validation experiments in future to substantiate the models performance in a clinical setting. User can access the tool named E-InfertilityTest as a standalone version on GitHub. Github Linkhttps://github.com/zglabDIB/einfertility.git

Dose-response modeling provides estimates of infectious and lethal doses, which can be used to inform control and prevention measures. Unfortunately, data from experimental challenge studies, which are needed to perform dose-response modeling, are often sparse. For example, non-human primate (NHP) challenge studies tend to have small samples sizes and little dose variation, often with only one or two dose levels per study. Thus, it is infeasible to apply traditional dose-response modeling approaches to data from single NHP studies. To address this challenge, we developed a mechanistic Bayesian model that aggregates and analyzes NHP pathogen load data across multiple studies. Our model links dose-infectivity to pathogen kinetics, which allows us to estimate the infectious dose and evaluate dose effects on within-host viral kinetics simultaneously. With this model, we obtained the first-ever ID50 estimate for SARS-CoV-1 in NHPs using data compiled from six NHP challenge studies. Our work demonstrates the value in reusing previous data from animal experiments. Our modeling framework can be applied to other pathogens, enabling robust dose-response inference when individual challenge studies are inconclusive.

Background and AimsHepatitis E virus (HEV) infections are a growing threat to global public health. To obtain an in-depth understanding of HEV infections in untreated and ribavirin-treated rats, we characterized the early HEV viral kinetics using rat HEV (rHEV) as a surrogate model and using two routes of virus inoculation: intravenous (I.V.) or oral infection. Approach and ResultsWe frequently collected feces, serum, and tissue samples up to 60 days after infection in both infection models to characterize the rHEV viral kinetics. A ~2-week delay in quantifiable RNA levels in feces was observed in the oral versus the I.V. infection model. Early rHEV viral kinetics in feces were found to be multiphasic and showed good concordance with those in the various tissue compartments studied. Comparison of the viral kinetics in these samples also revealed that the liver may serve as the initial site of rHEV replication, followed by replication in the intestine and spleen. While a dosage of 60 mg/kg/day ribavirin was found optimal to maintain rHEV RNA levels at (nearly) undetectable in feces, levels were detectable in the liver and increased both in feces and liver after treatment discontinuation. ConclusionsWe found that the two rHEV infection models share similar multiphasic viral kinetics with the liver as the main site of viral replication. Additionally, the rHEV RNA load in feces could be used as a reliable proxy for that in the liver, spleen, and intestine. We also show that ribavirin at 60 mg/kg/day was partially effective in preventing viral rebound. These findings may aid in exploring the correlation between the infection phases and antiviral efficacy, ultimately guiding therapeutic decisions.

AZD5582 (AZD) is a latency reversing agent used to support the "shock-and-kill" strategy in HIV-1 cure research. Previous studies in ART-suppressed rhesus macaques have shown that AZD can promote reactivation of latently infected cells, resulting in 2-3 log increases in on-ART viral load and significant reductions in SIV reservoir size over 5-10 doses. To quantify the impact of AZD on the reservoir, we developed an ensemble of mechanistic viral dynamic models and fit them to longitudinal plasma SIV RNA and SIV CA-DNA data from 23 macaques treated with AZD in combination with other therapies. The aggregate predictions of the model ensemble recapitulate the reactivation patterns observed in both SIV RNA and SIV CA-DNA and provide robust estimates of key parameters associated with reactivated cells. We found that AZD reactivates approximately 25% of cells in the latent reservoir per dose, with a mean reactivation duration of about 5-6 days. Of the reactivated cells, 60-79% are eventually cleared, while the remainder enter a state refractory to AZD stimulation before returning to latency. Because of this refractory state, each consecutive weekly dose reactivates about 28% fewer cells than the previous one, an effect that could be more pronounced if the refractory period substantially exceeds the interval between doses. However, the duration of this refractory state remains uncertain. Altogether, our results suggest that AZD-reactivated cells are effectively cleared. Future work should focus on improving LRAs that safely reactivate a larger fraction of the latent reservoir. Furthermore, designing experiments with varying dosing schedules can help better quantify the duration of refractoriness, which will be important for informing optimal treatment schedules and maximizing the effect of LRAs.

In the dynamic interplay between hosts and pathogens, hosts may produce a defense compound that acts as a toxin to deter pathogen attack. Conversely, pathogens may evolve to produce a counter-defense enzyme, neutralizing the hosts toxin. This evolutionary arms race incurs costs for both parties, prompting adaptations and strategic shifts. We conceptualize this interaction as an asymmetric game, with hosts and pathogens as players, and their potential responses - defense, counter-defense, or inaction - as their strategic options. In this scenario, if the pathogens counter-defense enzyme is entirely effective, then the hosts toxin is rendered obsolete. However, should the host cease toxin production, the pathogens enzyme becomes redundant, ironically reinstating the toxins utility. This interaction leads to potential red-queen cycles in defense and counter-defense strategies under certain conditions, or a balanced, optimal production of toxin and enzymes by hosts and parasites, respectively. To explore this, we introduce a game-theoretical model incorporating replicator dynamics to examine temporal shifts in strategy from active (counter-)defense to non-(counter-)defense and back. In addition, we analyze compromise strategies and interpret them as bet-hedging-like. We provide a deterministic illustration of how partial defense and counter-defense generate a fitness-buffering structure in unpredictable environments and increase the geometric mean fitness of the population. In conclusion, our analysis supports the notion of continuous periodic adjustments in strategies, notably in the levels of defensive and counter-defensive measures.

Adeno-associated virus (AAV) vectors are widely used in gene therapy, whereas low manufacturing efficiency and a large proportion of empty capsids are major obstacles. This study focused on the Yin Yang 1 (YY1) binding motif (YY1-motif) and investigated the effect of its presence or insertion at upstream of the Replicase (Rep)/Capsid Cap) gene on AAV vector production. We found that the YY1-motif incidentally presented in a Rep/Cap plasmid was associated with high vector production. We then designed several modified Rep/Cap (RC2) constructs. The YY1-motif insertion at the upstream of Rep/Cap gene increased vector yield in a repeat-number-dependent manner, and similar effects were not observed with other promoters insertion. Furthermore, the insertion of the YY1-motif reduced the amount of Cap protein per the same amount of full particle in supernatants on multiple serotypes, indicating the improvement in the empty/full capsid ratio. The YY1-motif insertion did not affect the AAV vector infectivity. These results denote that the YY1-motif has a universal regulatory function that optimizes the Rep/Cap expression balance, and simultaneously improves the production efficiency and full particle formation of AAV vectors. This finding could contribute to the development of highly efficient and high-quality AAV manufacturing processes.

Mycoplasma bovis is a minimal bacterium infecting cattle, which causes a wide variety of symptoms and is impacting dairy and beef producers worldwide. Part of the difficulty in research surrounding M. bovis, and other mycoplasmas, is the lack of efficient genome editing tools. As a proof of concept, we previously presented a transposon-based CRISPR-Base Editor system to introduce targeted mutations in M. bovis. In this work, the existing tool has been greatly improved: multi-loci targeting through addition of a second guide RNA; increased number of targetable loci by using an engineered Cas9 with AT-rich PAM specificity, and elimination of the CRISPR-Base Editor from the generated mutants through either transposon excision or use of a curable plasmid. We also propose a dedicated bioinformatic tool to identify target sequences in genes of a given genome. This software was applied to demonstrate the potential of our improved tools in M. bovis and other mycoplasmas of veterinary and human interest that currently lack genome editing methods.

This study examines the effects of rural out-migration and remittance inflows on food consumption outcomes among rural households in the Tigray region of Ethiopia. Utilizing household survey data collected from 521 rural households across three distinct Weredas (districts) (Tahtay Maichew, Kola Tembien, and Kilte-awlaelo). A Binary Probit model was employed to identify factors influencing migration decisions, while an Endogenous Switching Regression (ESR) model was used to estimate the impact of migration on food consumption outcomes while controlling for selection bias and unobserved heterogeneity. Food security was measured using the Food Consumption Score (FCS) and dietary diversity indicators. The empirical results reveal that severe food insecurity is widespread, with over 60% of all surveyed households falling into the "Poor" food consumption category. Descriptive baseline comparisons show that migration and remittance transfers marginally shift the raw average FCS upward from 23.86 to 25.48. However, this impact is profoundly nuanced: remittances serve as an immediate consumption-smoothing safety net but run parallel to a "labor-lost" constraint that reduces own-production capacities, forcing households to rely increasingly on market purchases for staple foods. The findings reveal that migration creates short-term labor shortages in agricultural production; however, remittance inflows substantially improve household food consumption frequencies, particularly for pulses, vegetables, and other nutrient-rich foods. After accounting for self-selection bias and unobserved traits, the rigorous ESR estimates indicate that migration increases the Food Consumption Score of participating households by an average Treatment Effect on the Treated (ATT) of 10.75 points, shifting them into more secure dietary tiers. Moreover, remittances help households mitigate the adverse effects of drought and other shocks by relaxing liquidity constraints and supporting both food purchases and agricultural investments. The study recommends establishing target food security safety nets for non-remittance households, promoting scale-appropriate labor-saving agricultural technologies, expanding traditional communal labor-sharing innovations, and boosting irrigation and agricultural input support programs to enhance rural food security and livelihood resilience.