Data-driven Computational Model Research Articles

Improving Anticancer Drug Selection and Prioritization via Neural Learning to Rank.

Personalized cancer treatment requires a thorough understanding of complex interactions between drugs and cancer cell lines in varying genetic and molecular contexts. To address this, high-throughput screening has been used to generate large-scale drug response data, facilitating data-driven computational models. Such models can capture complex drug-cell line interactions across various contexts in a fully data-driven manner. However, accurately prioritizing the most effective drugs for each cell line still remains a significant challenge. To address this, we developed multiple neural ranking approaches that leverage large-scale drug response data across multiple cell lines from diverse cancer types. Unlike existing approaches that primarily utilize regression and classification techniques for drug response prediction, we formulated the objective of drug selection and prioritization as a drug ranking problem. In this work, we proposed multiple pairwise and listwise neural ranking methods that learn latent representations of drugs and cell lines and then use those representations to score drugs in each cell line via a learnable scoring function. Specifically, we developed neural pairwise and listwise ranking methods, Pair-PushC and List-One on top of the existing methods, pLETORg and ListNet, respectively. Additionally, we proposed a novel listwise ranking method, List-All, that focuses on all the effective drugs instead of the top effective drug, unlike List-One. We also provide an exhaustive empirical evaluation with state-of-the-art regression and ranking baselines on large-scale data sets across multiple experimental settings. Our results demonstrate that our proposed ranking methods mostly outperform the best baselines with significant improvements of as much as 25.6% in terms of selecting truly effective drugs within the top 20 predicted drugs (i.e., hit@20) across 50% test cell lines. Furthermore, our analyses suggest that the learned latent spaces from our proposed methods demonstrate informative clustering structures and capture relevant underlying biological features. Moreover, our comprehensive evaluation provides a thorough and objective comparison of the performance of different methods (including our proposed ones).

Parallel ensemble Kalman method with total variation regularization for large-scale field inversion

Field inversion is often encountered in data-driven computational modeling to infer latent spatial–varying parameters from available observations. The ensemble Kalman method is emerging as a useful tool for solving field inversion problems due to its derivative-free merits. However, the method is computationally prohibitive for large-scale field inversion with high-dimensional observation data, which necessitates developing a practical efficient implementation strategy. In this work, we propose a parallel implementation of the ensemble Kalman method with total variation regularization for large-scale field inversion problems. It is achieved by partitioning the computational domain into non-overlapping subdomains and performing local ensemble Kalman updates at each subdomain parallelly. In doing so, the computational complexity of the ensemble-based inversion method is significantly reduced to the level of local subdomains. Further, the total variation regularization is employed to smoothen the physical field over the entire domain, which can reduce the inference discrepancy caused by missing covariances near subdomain interfaces. The capability of the proposed method is demonstrated in three field inversion problems with increasing complexity, i.e., the diffusion problem, the scalar transport problem and the Reynolds averaged Navier-Stokes closure problem. The numerical results show that the proposed method can significantly improve computational efficiency with satisfactory inference accuracy.

A general exposome factor explains individual differences in functional brain network topography and cognition in youth

Childhood environments are critical in shaping cognitive neurodevelopment. With the increasing availability of large-scale neuroimaging datasets with deep phenotyping of childhood environments, we can now build upon prior studies that have considered relationships between one or a handful of environmental and neuroimaging features at a time. Here, we characterize the combined effects of hundreds of inter-connected and co-occurring features of a child’s environment (“exposome”) and investigate associations with each child’s unique, multidimensional pattern of functional brain network organization (“functional topography”) and cognition. We apply data-driven computational models to measure the exposome and define personalized functional brain networks in pre-registered analyses. Across matched discovery (n=5139, 48.5% female) and replication (n=5137, 47.1% female) samples from the Adolescent Brain Cognitive Development study, the exposome was associated with current (ages 9–10) and future (ages 11–12) cognition. Changes in the exposome were also associated with changes in cognition after accounting for baseline scores. Cross-validated ridge regressions revealed that the exposome is reflected in functional topography and can predict performance across cognitive domains. Importantly, a single measure capturing a child’s exposome could more accurately and parsimoniously predict cognition than a wealth of personalized neuroimaging data, highlighting the importance of children’s complex, multidimensional environments in cognitive neurodevelopment.

PyCoM: a python library for large-scale analysis of residue-residue coevolution data.

Computational methods to detect correlated amino acid positions in proteins have become a valuable tool to predict intra- and inter-residue protein contacts, protein structures, and effects of mutation on protein stability and function. While there are many tools and webservers to compute coevolution scoring matrices, there is no central repository of alignments and coevolution matrices for large-scale studies and pattern detection leveraging on biological and structural annotations already available in UniProt. We present a Python library, PyCoM, which enables users to query and analyze coevolution matrices and sequence alignments of 457622 proteins, selected from UniProtKB/Swiss-Prot database (length ≤ 500 residues), from a precompiled coevolution matrix database (PyCoMdb). PyCoM facilitates the development of statistical analyses of residue coevolution patterns using filters on biological and structural annotations from UniProtKB/Swiss-Prot, with simple access to PyCoMdb for both novice and advanced users, supporting Jupyter Notebooks, Python scripts, and a web API access. The resource is open source and will help in generating data-driven computational models and methods to study and understand protein structures, stability, function, and design. PyCoM code is freely available from https://github.com/scdantu/pycom and PyCoMdb and the Jupyter Notebook tutorials are freely available from https://pycom.brunel.ac.uk.

Integrating inverse reinforcement learning into data-driven mechanistic computational models: a novel paradigm to decode cancer cell heterogeneity

Cellular heterogeneity is a ubiquitous aspect of biology and a major obstacle to successful cancer treatment. Several techniques have emerged to quantify heterogeneity in live cells along axes including cellular migration, morphology, growth, and signaling. Crucially, these studies reveal that cellular heterogeneity is not a result of randomness or a failure in cellular control systems, but instead is a predictable aspect of multicellular systems. We hypothesize that individual cells in complex tissues can behave as reward-maximizing agents and that differences in reward perception can explain heterogeneity. In this perspective, we introduce inverse reinforcement learning as a novel approach for analyzing cellular heterogeneity. We briefly detail experimental approaches for measuring cellular heterogeneity over time and how these experiments can generate datasets consisting of cellular states and actions. Next, we show how inverse reinforcement learning can be applied to these datasets to infer how individual cells choose different actions based on heterogeneous states. Finally, we introduce potential applications of inverse reinforcement learning to three cell biology problems. Overall, we expect inverse reinforcement learning to reveal why cells behave heterogeneously and enable identification of novel treatments based on this new understanding.

Computational pathology model to assess acute and chronic transformations of the tubulointerstitial compartment in renal allograft biopsies

Managing patients with kidney allografts largely depends on biopsy diagnosis which is based on semiquantitative assessments of rejection features and extent of acute and chronic changes within the renal parenchyma. Current methods lack reproducibility while digital image data-driven computational models enable comprehensive and quantitative assays. In this study we aimed to develop a computational method for automated assessment of histopathology transformations within the tubulointerstitial compartment of the renal cortex. Whole slide images of modified Picrosirius red-stained biopsy slides were used for the training (n = 852) and both internal (n = 172) and external (n = 94) tests datasets. The pipeline utilizes deep learning segmentations of renal tubules, interstitium, and peritubular capillaries from which morphometry features were extracted. Seven indicators were selected for exploring the intrinsic spatial interactions within the tubulointerstitial compartment. A principal component analysis revealed two independent factors which can be interpreted as representing chronic and acute tubulointerstitial injury. A K-means clustering classified biopsies according to potential phenotypes of combined acute and chronic transformations of various degrees. We conclude that multivariate analyses of tubulointerstitial morphometry transformations enable extraction of and quantification of acute and chronic components of injury. The method is developed for renal allograft biopsies; however, the principle can be applied more broadly for kidney pathology assessment.

Promote to protect: data-driven computational model of peer influence for vaccine perception

Vaccine hesitancy and acceptance, driven by social influence, is usually explored by most researchers using exhaustive survey-based studies, which investigate public preferences, fundamental values, beliefs, barriers, and drivers through closed or open-ended questionnaires. Commonly used simple statistical tools do not do justice to the richness of this data. Considering the gradual development of vaccine acceptance in a society driven by multiple local/global factors as a compartmental contagion process, we propose a novel methodology where drivers and barriers of these dynamics are detected from survey participants’ responses, instead of heuristic arguments. Applying rigorous natural language processing analysis to the survey responses of participants from India, who are from various socio-demographics, education, and perceptions, we identify and categorize the most important factors as well as interactions among people of different perspectives on COVID-19 vaccines. With a goal to achieve improvement in vaccine perception, we also analyze the resultant behavioral transitions through platforms of unsupervised machine learning and natural language processing to derive a compartmental contagion model from the data. Analysis of the model shows that positive peer influence plays a very important role and causes a bifurcation in the system that reflects threshold-sensitive dynamics.

Deep generative spatiotemporal learning for integrating fracture mechanics in composite materials: inverse design, discovery, and optimization.

The trade-off between strength and toughness presents a fundamental challenge in engineering material design. Composite materials (CMs) can strategically arrange different materials to enhance both strength and toughness by optimizing the distribution of loads and increasing resistance to crack propagation. However, current data-driven computational modeling approaches for CM configuration optimization suffer from limitations of "substantial computational cost" and "poor predictive power over extrapolation spaces", making it difficult to integrate with global optimization algorithms, and ultimately limiting the discovery of materials with optimal tradeoffs. As a breakthrough, we propose a data-driven design framework with a multi-task DL architecture capable of accurately predicting local fields' spatiotemporal behavior, including stress evolution and crack propagation, alongside homogenized mechanical properties. Our model, trained on datasets generated from crack phase fields simulations of random configurations, demonstrated exceptional predictive performance even for unseen configurations with well organized patterns exploiting nature-inspired morphological features. Importantly, solely from composite material (CM) configurations, our model effectively predicts long-term spatiotemporal fields with an accuracy comparable to FEM but with a substantial reduction in computational time. By coupling the model's predictive power with genetic optimization algorithms, we demonstrated the framework's applicability in two representative inverse design tasks: devising CM configurations with mechanical properties beyond the training set and guiding desired crack pattern formation. Our research highlights the potential of artificial intelligence as a feasible alternative to conventional computational approaches for straightforward configurational and structural optimization.

Computational multigene interactions in virus growth and infection spread.

Viruses persist in nature owing to their extreme genetic heterogeneity and large population sizes, which enable them to evade host immune defenses, escape antiviral drugs, and adapt to new hosts. The persistence of viruses is challenging to study because mutations affect multiple virus genes, interactions among genes in their impacts on virus growth are seldom known, and measures of viral fitness are yet to be standardized. To address these challenges, we employed a data-driven computational model of cell infection by a virus. The infection model accounted for the kinetics of viral gene expression, functional gene-gene interactions, genome replication, and allocation of host cellular resources to produce progeny of vesicular stomatitis virus, a prototype RNA virus. We used this model to computationally probe how interactions among genes carrying up to eleven deleterious mutations affect different measures of virus fitness: single-cycle growth yields and multicycle rates of infection spread. Individual mutations were implemented by perturbing biophysical parameters associated with individual gene functions of the wild-type model. Our analysis revealed synergistic epistasis among deleterious mutations in their effects on virus yield; so adverse effects of single deleterious mutations were amplified by interaction. For the same mutations, multicycle infection spread indicated weak or negligible epistasis, where single mutations act alone in their effects on infection spread. These results were robust to simulation in high- and low-host resource environments. Our work highlights how different types and magnitudes of epistasis can arise for genetically identical virus variants, depending on the fitness measure. More broadly, gene-gene interactions can differently affect how viruses grow and spread.

A data-driven computational model for obesity-driven diabetes onset and remission through weight loss

Obesity is a major risk factor for the development of type 2 diabetes (T2D), where a sustained weight loss may result in T2D remission in individuals with obesity. To design effective and feasible intervention strategies to prevent or reverse T2D, it is imperative to study the progression of T2D and remission together. Unfortunately, this is not possible through experimental and observational studies. To address this issue, we introduce a data-driven computational model and use human data to investigate the progression of T2D with obesity and remission through weight loss on the same timeline. We identify thresholds for the emergence of T2D and necessary conditions for remission. We explain why remission is only possible within a window of opportunity and the way that window depends on the progression history of T2D, individual's metabolic state, and calorie restrictions. These findings can help to optimize therapeutic intervention strategies for T2D prevention or treatment.

A data-driven computational model for engineered cardiac microtissues

Engineered heart tissues (EHTs) present a potential solution to some of the current challenges in the treatment of heart disease; however, the development of mature, adult-like cardiac tissues remains elusive. Mechanical stimuli have been observed to improve whole-tissue function and cardiomyocyte (CM) maturation, although our ability to fully utilize these mechanisms is hampered, in part, by our incomplete understanding of the mechanobiology of EHTs. In this work, we leverage experimental data, produced by a mechanically tunable experimental setup, to introduce a tissue-specific computational modeling pipeline of EHTs. Our new modeling pipeline generates simulated, image-based EHTs, capturing ECM and myofibrillar structure as well as functional parameters estimated directly from experimental data. This approach enables the unique estimation of EHT function by data-based estimation of CM active stresses. We use this experimental and modeling pipeline to study different mechanical environments, where we contrast the force output of the tissue with the computed active stress of CMs. We show that the significant differences in measured experimental forces can largely be explained by the levels of myofibril formation achieved by the CMs in the distinct mechanical environments, with active stress showing more muted variations across conditions. The presented model also enables us to dissect the relative contributions of myofibrils and extracellular matrix to tissue force output, a task difficult to address experimentally. These results highlight the importance of tissue-specific modeling to augment EHT experiments, providing deeper insights into the mechanobiology driving EHT function. Statement of significanceEngineered heart tissues (EHTs) have the potential to revolutionize the way heart disease is treated. However, developing mature cardiomyocytes (CM) in these tissues remains a challenge due, in part, to our incomplete understanding of the fundamental biomechanical mechanisms that drive EHT development. This work integrates the experimental data of an EHT platform developed to study the influence of mechanics in CM maturation with computational biomechanical models. This approach is used to augment conclusions obtained in-vitro - by measuring quantities such as cell stress and strain - and to dissect the relevance of each component in the whole tissue performance. Our results show how a combination of specialized in-silico and in-vitro approaches can help us better understand the mechanobiology of EHTs.

Trustworthiness judgments without the halo effect: A data-driven computational modeling approach.

Trustworthy-looking faces are also perceived as more attractive, but are there other meaningful cues that contribute to perceived trustworthiness? Using data-driven models, we identify these cues after removing attractiveness cues. In Experiment 1, we show that both judgments of trustworthiness and attractiveness of faces manipulated by a model of perceived trustworthiness change in the same direction. To control for the effect of attractiveness, we build two new models of perceived trustworthiness: a subtraction model, which forces the perceived attractiveness and trustworthiness to be negatively correlated (Experiment 2), and an orthogonal model, which reduces their correlation (Experiment 3). In both experiments, faces manipulated to appear more trustworthy were indeed perceived to be more trustworthy, but not more attractive. Importantly, in both experiments, these faces were also perceived as more approachable and with more positive expressions, as indicated by both judgments and machine learning algorithms. The current studies show that the visual cues used for trustworthiness and attractiveness judgments can be separated, and that apparent approachability and facial emotion are driving trustworthiness judgments and possibly general valence evaluation.

Analysis of Network Models with Neuron-Astrocyte Interactions

Neural networks, composed of many neurons and governed by complex interactions between them, are a widely accepted formalism for modeling and exploring global dynamics and emergent properties in brain systems. In the past decades, experimental evidence of computationally relevant neuron-astrocyte interactions, as well as the astrocytic modulation of global neural dynamics, have accumulated. These findings motivated advances in computational glioscience and inspired several models integrating mechanisms of neuron-astrocyte interactions into the standard neural network formalism. These models were developed to study, for example, synchronization, information transfer, synaptic plasticity, and hyperexcitability, as well as classification tasks and hardware implementations. We here focus on network models of at least two neurons interacting bidirectionally with at least two astrocytes that include explicitly modeled astrocytic calcium dynamics. In this study, we analyze the evolution of these models and the biophysical, biochemical, cellular, and network mechanisms used to construct them. Based on our analysis, we propose how to systematically describe and categorize interaction schemes between cells in neuron-astrocyte networks. We additionally study the models in view of the existing experimental data and present future perspectives. Our analysis is an important first step towards understanding astrocytic contribution to brain functions. However, more advances are needed to collect comprehensive data about astrocyte morphology and physiology in vivo and to better integrate them in data-driven computational models. Broadening the discussion about theoretical approaches and expanding the computational tools is necessary to better understand astrocytes’ roles in brain functions.

A Deep-Learning-Based Multi-Modal Sensor Fusion Approach for Detection of Equipment Faults

Condition monitoring is a part of the predictive maintenance approach applied to detect and prevent unexpected equipment failures by monitoring machine conditions. Early detection of equipment failures in industrial systems can greatly reduce scrap and financial losses. Developed sensor data acquisition technologies allow for digitally generating and storing many types of sensor data. Data-driven computational models allow the extraction of information about the machine’s state from acquired sensor data. The outstanding generalization capabilities of deep learning models have enabled them to play a significant role as a data-driven computational fault model in equipment condition monitoring. A challenge of fault detection applications is that single-sensor data can be insufficient in performance to detect equipment anomalies. Furthermore, data in different domains can reveal more prominent features depending on the fault type, but may not always be obvious. To address this issue, this paper proposes a multi-modal sensor fusion-based deep learning model to detect equipment faults by fusing information not only from different sensors but also from different signal domains. The effectiveness of the model’s fault detection capability is shown by utilizing the most commonly encountered equipment types in the industry, such as electric motors. Two different sensor types’ raw time domain and frequency domain data are utilized. The raw data from the vibration and current sensors are transformed into time-frequency images using short-time Fourier transform (STFT). Then, time-frequency images and raw time series data were supplied to the designed deep learning model to detect failures. The results showed that the fusion of multi-modal sensor data using the proposed model can be advantageous in equipment fault detection.

Computational models of neurotransmission at cerebellar synapses unveil the impact on network computation.

The neuroscientific field benefits from the conjoint evolution of experimental and computational techniques, allowing for the reconstruction and simulation of complex models of neurons and synapses. Chemical synapses are characterized by presynaptic vesicle cycling, neurotransmitter diffusion, and postsynaptic receptor activation, which eventually lead to postsynaptic currents and subsequent membrane potential changes. These mechanisms have been accurately modeled for different synapses and receptor types (AMPA, NMDA, and GABA) of the cerebellar cortical network, allowing simulation of their impact on computation. Of special relevance is short-term synaptic plasticity, which generates spatiotemporal filtering in local microcircuits and controls burst transmission and information flow through the network. Here, we present how data-driven computational models recapitulate the properties of neurotransmission at cerebellar synapses. The simulation of microcircuit models is starting to reveal how diverse synaptic mechanisms shape the spatiotemporal profiles of circuit activity and computation.

Synaptic Plasticity Is Predicted by Spatiotemporal Firing Rate Patterns and Robust to In Vivo-like Variability.

Synaptic plasticity, the experience-induced change in connections between neurons, underlies learning and memory in the brain. Most of our understanding of synaptic plasticity derives from in vitro experiments with precisely repeated stimulus patterns; however, neurons exhibit significant variability in vivo during repeated experiences. Further, the spatial pattern of synaptic inputs to the dendritic tree influences synaptic plasticity, yet is not considered in most synaptic plasticity rules. Here, we investigate how spatiotemporal synaptic input patterns produce plasticity with in vivo-like conditions using a data-driven computational model with a plasticity rule based on calcium dynamics. Using in vivo spike train recordings as inputs to different size clusters of spines, we show that plasticity is strongly robust to trial-to-trial variability of spike timing. In addition, we derive general synaptic plasticity rules describing how spatiotemporal patterns of synaptic inputs control the magnitude and direction of plasticity. Synapses that strongly potentiated have greater firing rates and calcium concentration later in the trial, whereas strongly depressing synapses have hiring firing rates early in the trial. The neighboring synaptic activity influences the direction and magnitude of synaptic plasticity, with small clusters of spines producing the greatest increase in synaptic strength. Together, our results reveal that calcium dynamics can unify diverse plasticity rules and reveal how spatiotemporal firing rate patterns control synaptic plasticity.

An expert judgment model to predict early stages of the COVID-19 pandemic in the United States.

From February to May 2020, experts in the modeling of infectious disease provided quantitative predictions and estimates of trends in the emerging COVID-19 pandemic in a series of 13 surveys. Data on existing transmission patterns were sparse when the pandemic began, but experts synthesized information available to them to provide quantitative, judgment-based assessments of the current and future state of the pandemic. We aggregated expert predictions into a single "linear pool" by taking an equally weighted average of their probabilistic statements. At a time when few computational models made public estimates or predictions about the pandemic, expert judgment provided (a) falsifiable predictions of short- and long-term pandemic outcomes related to reported COVID-19 cases, hospitalizations, and deaths, (b) estimates of latent viral transmission, and (c) counterfactual assessments of pandemic trajectories under different scenarios. The linear pool approach of aggregating expert predictions provided more consistently accurate predictions than any individual expert, although the predictive accuracy of a linear pool rarely provided the most accurate prediction. This work highlights the importance that an expert linear pool could play in flexibly assessing a wide array of risks early in future emerging outbreaks, especially in settings where available data cannot yet support data-driven computational modeling.

Data-driven computational models of ventricular-arterial hemodynamics in pediatric pulmonary arterial hypertension.

Pulmonary arterial hypertension (PAH) is a complex disease involving increased resistance in the pulmonary arteries and subsequent right ventricular (RV) remodeling. Ventricular-arterial interactions are fundamental to PAH pathophysiology but are rarely captured in computational models. It is important to identify metrics that capture and quantify these interactions to inform our understanding of this disease as well as potentially facilitate patient stratification. Towards this end, we developed and calibrated two multi-scale high-resolution closed-loop computational models using open-source software: a high-resolution arterial model implemented using CRIMSON, and a high-resolution ventricular model implemented using FEniCS. Models were constructed with clinical data including non-invasive imaging and invasive hemodynamic measurements from a cohort of pediatric PAH patients. A contribution of this work is the discussion of inconsistencies in anatomical and hemodynamic data routinely acquired in PAH patients. We proposed and implemented strategies to mitigate these inconsistencies, and subsequently use this data to inform and calibrate computational models of the ventricles and large arteries. Computational models based on adjusted clinical data were calibrated until the simulated results for the high-resolution arterial models matched within 10% of adjusted data consisting of pressure and flow, whereas the high-resolution ventricular models were calibrated until simulation results matched adjusted data of volume and pressure waveforms within 10%. A statistical analysis was performed to correlate numerous data-derived and model-derived metrics with clinically assessed disease severity. Several model-derived metrics were strongly correlated with clinically assessed disease severity, suggesting that computational models may aid in assessing PAH severity.

Optimal anti-amyloid-beta therapy for Alzheimer's disease via a personalized mathematical model.

With the recent approval by the FDA of the first disease-modifying drug for Alzheimer's Disease (AD), personalized medicine will be increasingly important for appropriate management and counseling of patients with AD and those at risk. The growing availability of clinical biomarker data and data-driven computational modeling techniques provide an opportunity for new approaches to individualized AD therapeutic planning. In this paper, we develop a new mathematical model, based on AD cognitive, cerebrospinal fluid (CSF) and MRI biomarkers, to provide a personalized optimal treatment plan for individuals. This model is parameterized by biomarker data from the AD Neuroimaging Initiative (ADNI) cohort, a large multi-institutional database monitoring the natural history of subjects with AD and mild cognitive impairment (MCI). Optimal control theory is used to incorporate time-varying treatment controls and side-effects into the model, based on recent clinical trial data, to provide a personalized treatment regimen with anti-amyloid-beta therapy. In-silico treatment studies were conducted on the approved treatment, aducanumab, as well as on another promising anti-amyloid-beta therapy under evaluation, donanemab. Clinical trial simulations were conducted over both short-term (78 weeks) and long-term (10 years) periods with low-dose (6 mg/kg) and high-dose (10 mg/kg) regimens for aducanumab, and a single-dose regimen (1400 mg) for donanemab. Results confirm those of actual clinical trials showing a large and sustained effect of both aducanumab and donanemab on amyloid beta clearance. The effect on slowing cognitive decline was modest for both treatments, but greater for donanemab. This optimal treatment computational modeling framework can be applied to other single and combination treatments for both prediction and optimization, as well as incorporate new clinical trial data as it becomes available.

Cell-type-specific population dynamics of diverse reward computations

Computational analysis of cellular activity has developed largely independently of modern transcriptomic cell typology, but integrating these approaches may be essential for full insight into cellular-level mechanisms underlying brain function and dysfunction. Applying this approach to the habenula (a structure with diverse, intermingled molecular, anatomical, and computational features), we identified encoding of reward-predictive cues and reward outcomes in distinct genetically defined neural populations, including TH+ cells and Tac1+ cells. Data from genetically targeted recordings were used to train an optimized nonlinear dynamical systems model and revealed activity dynamics consistent with a line attractor. High-density, cell-type-specific electrophysiological recordings and optogenetic perturbation provided supporting evidence for this model. Reverse-engineering predicted how Tac1+ cells might integrate reward history, which was complemented by invivo experimentation. This integrated approach describes a process by which data-driven computational models of population activity can generate and frame actionable hypotheses for cell-type-specific investigation in biological systems.