Biology subjects

Ai, L.

Publications and source records attributed to Ai, L..

At least 19 recordsLinked to original sources

bioRxiv2026

Turning Domain Expertise into Multi-Dimensional Evaluation of Biomedical AI with Karenina

Language models and agents are increasingly used in biomedicine, but current benchmarks reward correct answers even when the underlying reasoning is flawed. Here we introduce Karenina, an open-source framework that turns expert knowledge into multi-dimensional evaluations of questions, conversations and autonomous agents. Illustrated in Question-Answer pairs, multi-turn conversations and autonomous data-analysis, these dimensions together moves evaluation beyond scoring, enabling trustworthy decision-making with AI in biomedicine.

bioinformatics↗

bioRxiv2026

Expanded Proteome Coverage Powered by Advanced Ion Processing Enables Deep Single-Cell Drug Response Subtyping in Human Stem Cell Derived Cardiomyocytes

Single-cell proteomics (SCP) enables the study of cellular heterogeneity at the functional level but remains limited by incomplete proteome coverage and high data missingness. Here, we present an enhanced label-free SCP workflow that leverages the timsUltra AIP mass spectrometry platform equipped with the Athena Ion Processor (AIP). Across a controlled dilution series of human induced pluripotent stem cell-derived cardiomyocytes (iPSC-CMs), AIP-enabled acquisition consistently increased proteome depth and detection consistency across cells at all input levels. In single iPSC-CMs, the timsUltra AIP quantified up to 3,858 protein groups, averaging [~]1,300 proteins per cell, enabling robust proteome-level classification of cardiomyocyte subtypes. Using a reference-based protein classifier, cells were stratified into mature cardiomyocytes and less differentiated cell states, revealing substantial baseline heterogeneity. Importantly, increased single-cell sensitivity translated directly into biological insight, as approximately 30% of differentially expressed proteins associated with subtype-specific drug responses were detected exclusively by timsUltra AIP. Application of this workflow to PR-364 (a mitophagy boosting drug) dose-response experiment uncovered distinct, subtype-dependent pathway adaptations. Mature cardiomyocytes exhibited dose-dependent increases in mitochondrial and metabolic pathway activity, while immature cells showed enrichment of cytoskeletal and developmental programs. These effects were partially obscured in simulated bulk analyses, highlighting the value of single-cell resolution. Together, these results demonstrate that improved fragment ion transmission and utilization translate directly into enhanced biological insight, enabling more comprehensive and functionally relevant single-cell proteomics.

molecular biology↗

bioRxiv2025

CerS2 is a druggable target in triple-negative breast cancer

Triple-negative breast cancer (TNBC) remains a significant clinical challenge due to its aggressive nature and lack of effective targeted therapies. The enzyme ceramide synthase 2 (CerS2), which synthesizes pro-apoptotic very long-chain ceramides (VLCCs), represents a promising therapeutic target. Here, we identify and characterize DH20931, a novel, first-in-class small-molecule agonist of CerS2. We demonstrate that DH20931 directly activates CerS2 with nanomolar potency, leading to significant VLCC accumulation in breast cancer cells. This lipotoxic event induces endoplasmic reticulum (ER) stress and triggers apoptosis via the canonical ATF4/CHOP/PUMA signaling pathway. Mechanistically, we uncover a novel interaction between CerS2 and the ER calcium channel, Inositol 1,4,5-trisphosphate receptor 1 (IP3R1). We demonstrate that DH20931 promotes this interaction, enhancing ER-mitochondria proximity and facilitating a CerS2-dependent flux of calcium (Ca{superscript 2}) from the ER into mitochondria. This subsequent mitochondrial Ca{superscript 2} overload serves as a critical trigger for apoptosis. In preclinical evaluations, DH20931 potently inhibited the growth of TNBC cells in 2D and 3D cultures and significantly suppressed tumor progression in orthotopic and patient-derived xenograft (PDX) models, all while exhibiting a favorable safety profile. Our findings validate CerS2 as a druggable target in TNBC and establish a novel therapeutic strategy that leverages a coordinated attack on cancer cells through ER stress and calcium-mediated mitochondrial dysfunction. Graphical Abstract O_FIG O_LINKSMALLFIG WIDTH=200 HEIGHT=100 SRC="FIGDIR/small/670525v1_ufig1.gif" ALT="Figure 1"> View larger version (33K): org.highwire.dtl.DTLVardef@2db377org.highwire.dtl.DTLVardef@6175d4org.highwire.dtl.DTLVardef@7edb8corg.highwire.dtl.DTLVardef@80042e_HPS_FORMAT_FIGEXP M_FIG C_FIG HighlightsO_LIDH20931 is a first-in-class, potent agonist of Ceramide Synthase 2 (CerS2). C_LIO_LICerS2 activation induces ER stress and engages the ATF4/CHOP/PUMA apoptotic pathway. C_LIO_LIDH20931 promotes a novel CerS2-IP3R1 interaction, causing lethal mitochondrial calcium overload. C_LIO_LITargeting CerS2 shows significant preclinical efficacy against triple-negative breast cancer. C_LI

cancer biology↗

bioRxiv2025

Reproducible Brain Charts: An open data resource for mapping brain development and its associations with mental health

Major mental disorders are increasingly understood as disorders of brain development. Large and heterogeneous samples are required to define generalizable links between brain development and psychopathology. To this end, we introduce the Reproducible Brain Charts (RBC), an open data resource that integrates data from 5 large studies of brain development in youth from three continents (N=6,346; 45% Female). Confirmatory bifactor models were used to create harmonized psychiatric phenotypes that capture major dimensions of psychopathology. Following rigorous quality assurance, neuroimaging data were carefully curated and processed using consistent pipelines in a reproducible manner with DataLad, the Configurable Pipeline for the Analysis of Connectomes (C-PAC), and FreeSurfer. Initial analyses of RBC data emphasize the benefit of careful quality assurance and data harmonization in delineating developmental effects and associations with psychopathology. Critically, all RBC data - including harmonized psychiatric phenotypes, unprocessed images, and fully processed imaging derivatives - are openly shared without a data use agreement via the International Neuroimaging Data-sharing Initiative. Together, RBC facilitates large-scale, reproducible, and generalizable research in developmental and psychiatric neuroscience.

neuroscience↗

bioRxiv2024

A CRISPR/SpCas9M-reporting system for efficient and rapid genome editing in Caulobacter crescentus

As members of the -proteobacteria group, Caulobacter crescentus and its relatives are known for their asymmetric life cycle and comprehensive applications in gene delivery, agricultural biotechnology, and the production of high-value compounds. However, genetic manipulations of these bacteria are often time-consuming and labor-intensive due to the lack of efficient genome editing tools. Here, we report a practical CRISPR/SpCas9M-reporting system that overcomes the limitations of SpCas9 expression, enabling efficient, markerless, and rapid genome editing in C. crescentus. As a demonstration, we successfully knocked out two genes encoding the scaffold proteins, achieving apparent editing efficiencies up to 80%. Key components, including the Cas protein, Cas inducer, sgRNA, homologous arms, and reporter, were systematically analyzed and optimized to enhance the editing efficiency or decrease the cell lethality. A nearly zero off-target ratio was observed after the curing of the editor plasmid in editing strains. Furthermore, we applied the CRISPR/SpCas9M-reporting system to two C. crescentus relatives, Agrobacterium fabrum and Sinorhizobium meliloti, establishing it as an efficient and reliable editing strategy. We anticipate that this system could be applied to other hard-to-edit organisms, accelerating both basic and applied research in -proteobacteria.

bioengineering↗

bioRxiv2024

Distributed Collaboration for Data, Analysis Pipelines, and Results in Single-Cell Omics

Single-cell omics data analysis pipelines are complicated to design and difficult to share or reproduce. We describe a web platform that enables no-code analysis pipeline design, simple computing via the Open Science Grid, and sharing of entire data analysis pipelines, their input data, and interactive results. We expect this platform to increase the accessibility and reproducibility of single-cell omics.

cell biology↗

bioRxiv2023

A Novel HIF-2α/ARNT Signaling Pathway Protects Against Microvascular Dysfunction and heart failure After Myocardial Infarction

Myocardial infarction (MI) significantly compromises the integrity of the cardiac microvascular endothelial barrier, leading to enhanced leakage and inflammation that contribute to the progression of heart failure. While HIF2 is highly expressed in cardiac endothelial cells (ECs) under hypoxic conditions, its role in regulating microvascular endothelial barrier function during MI is not well understood. In this study, we utilized mice with a cardiac-specific deletion of HIF2, generated through an inducible Cre (Cdh5Cre-ERT2) recombinase system. These mice exhibited no apparent phenotype under normal conditions. However, following left anterior descending (LAD) artery ligation-induced MI, they showed increased mortality associated with enhanced cardiac vascular leakage, inflammation, worsened cardiac function, and exacerbated heart remodeling. These outcomes suggest a protective role for endothelial HIF2 in response to cardiac ischemia. Parallel investigations in human cardiac microvascular endothelial cells (CMVECs) revealed that loss of ecHif2 led to diminished endothelial barrier function, characterized by reduced tight-junction protein levels and increased cell death, along with elevated expression of IL6 and other inflammatory markers. These effects were substantially reversed by overexpressing ARNT, a critical dimerization partner for HIF2 during hypoxia. Additionally, ARNT deletion also led to increased CMVEC permeability. Interestingly, ARNT, rather than HIF2 itself, directly binds to the IL6 promoter to suppress IL6 expression. Our findings demonstrate the critical role of endothelial HIF2 in response to MI and identify the HIF2/ARNT axis as a transcriptional repressor, offering novel insights for developing therapeutic strategies against heart failure following MI.

cell biology↗

bioRxiv2022

High Field Asymmetric Waveform Ion Mobility Spectrometry- Mass Spectrometry to Enhance Cardiac Muscle Proteome Coverage

Heart tissue sample preparation for mass spectrometry (MS) analysis that includes pre-fractionation reduces the cellular protein dynamic range and increases the relative abundance of non-sarcomeric proteins. We previously described "IN-Sequence" (IN-Seq) where heart tissue lysate is sequentially partitioned into three subcellular fractions to increase the proteome coverage than a single direct tissue analysis by mass spectrometry. Here, we report an adaptation of the high-field asymmetric ion mobility spectrometry (FAIMS) coupled to mass spectrometry, and the establishment of a simple one step sample preparation coupled with gas-phase fractionation. FAIMS approach substantially reduces manual sample handling, significantly shortens MS instrument processing time, and produces unique protein identification and quantification approximating the commonly used IN-Seq method in for less time requirement.

molecular biology↗

bioRxiv2022

An open, analysis-ready, and quality controlled resource for pediatric brain white-matter research

We created resources to facilitate research on the role of human brain microstructure in the development of mental health disorders, based on openly-available diffusion MRI (dMRI) data from the Healthy Brain Network (HBN) study. First, we curated the HBN dMRI data (N=2747) into the Brain Imaging Data Structure and preprocessed it according to best-practices, including denoising and correcting for motion effects, susceptibility-related distortions, and eddy currents. Preprocessed, analysis-ready data was made openly available. Data quality plays a key role in the analysis of dMRI, and we provide automated quality control (QC) scores for every scan, as part of the data release. To scale QC to this large dataset, we trained a neural network through the combination of a small data subset scored by experts and a larger set scored by community scientists. The network performs QC highly concordant with that of experts on a held out set (ROC-AUC = 0.947). A further analysis of the neural network demonstrates that it relies on image features with relevance to QC. Altogether, this work both delivers a resource for transdiagnostic research in brain connectivity and pediatric mental health and serves as a novel tool for automated QC of large datasets.

neuroscience↗

bioRxiv2022

Lyophilized mRNA-lipid nanoparticles vaccine with long-term stability and high antigenicity against SARS-CoV-2

Advanced mRNA vaccines play vital roles against SARS-CoV-2. However, due to their poor stability, most current mRNA delivery platforms need to be stored at -20{degrees}C or -70{degrees}C, which severely limits their distribution. Herein, we present lyophilized SARS-CoV-2 mRNA-lipid nanoparticle vaccines, which can be stored at room temperature with long-term thermostability. In the in vivo Delta virus challenge experiment, lyophilized Delta variant mRNA vaccine successfully protected mice from infection and cleared the virus. Lyophilized omicron mRNA vaccine enabled to elicit both potent humoral and cellular immunity. In booster immunization experiments in mice and old monkeys, lyophilized omicron mRNA vaccine could effectively increase the titers of neutralizing antibodies against wild-type coronavirus and omicron variants. In humans, lyophilized omicron mRNA vaccine as a booster shot could also engender excellent immunity and had less severe adverse events. This lyophilization platform overcomes the instability of mRNA vaccines without affecting their bioactivity, and significantly improved their accessibility, particularly in remote regions.

immunology↗

bioRxiv2021

Moving Beyond Processing and Analysis-Related Variation in Neuroscience

When fields lack consensus standard methods and accessible ground truths, reproducibility can be more of an ideal than a reality. Such has been the case for functional neuroimaging, where there exists a sprawling space of tools and processing pipelines. We provide a critical evaluation of the impact of differences across five independently developed minimal preprocessing pipelines for functional MRI. We show that even when handling identical data, inter-pipeline agreement was only moderate, critically shedding light on a factor that limits cross-study reproducibility. We show that low inter-pipeline agreement mainly becomes appreciable when the reliability of the underlying data is high, which is increasingly the case as the field progresses. Crucially, we show that when inter-pipeline agreement is compromised, so too are the consistency of insights from brainwide association studies. We highlight the importance of comparing analytic configurations, as both widely discussed and commonly overlooked decisions can lead to marked variation.

neuroscience↗

bioRxiv2021

An NCCN-IPI based immune-related gene prognostic model for diffuse large B-cell lymphoma

BackgroundInternational Prognostic Index (IPI) was widely used to better discriminate prognosis of patients with diffuse large B-cell lymphoma (DLBCL). However, there is a significant need to identify novel valuable biomarkers in the context of targeted therapies, such as immune checkpoint blockade (ICB) therapy. MethodsGene expression data and clinical information of DLBCL were obtained from The Cancer Genome Atlas (TCGA) and Gene Expression Omnibus (GEO) datasets. 371 immune-related hub genes in DLBCL patients with different IPI levels were identified by weighted gene co-expression network analysis (WGCNA), and 8 genes were selected to construct an IPI-based immune prognostic model (IPI-IPM). Afterward, the genetic, somatic mutational and molecular profiles of IPI-IPM subgroups were analyzed, as well as the potential clinical response of ICB in different IPI-IPM subgroups. ResultsThe IPI-IPM was constructed based on the expression of CMBL, TLCD3B, SYNDIG1, ESM1, EPHA3, HUNK, PTX3 and IL12A, where high-risk patients had shorter overall survival (OS) than low-risk patients, consistent with the results in the GEO cohorts. The comprehensive results showed that high IPI-IPM risk scores were correlated with immune-related signaling pathways, high KMT2D and CD79B mutation rates, as well as up-regulation of inhibitory immune checkpoints including PD-L1, BTLA and SIGLEC7, indicating more potential response to ICB therapy. ConclusionThe IPI-IPM has independent prognostic significance for DLBCL patients, which provides an immunological perspective to elucidate the mechanisms on tumor progression, also sheds a light on developing immunotherapy for DLBCL.

bioinformatics↗

bioRxiv2020

Association of CD44-/CD24- Breast Cancer Cells with Late Stage Tumor Recurrence

Tumor metastasis remains the main cause of breast cancer-related deaths, especially the later breast cancer distant metastasis. This study assessed CD44-/CD24- tumor cells in 576 tissue specimens for associations with clinicopathological features and metastasis and then investigated the underlying molecular events. The data showed that level of CD44-/CD24- cells was associated with later postoperative distant tumor metastasis. Furthermore, CD44-/CD24- triple negative cells could spontaneously convert into CD44+/CD24- cancer stem cells (CSCs) with properties similar to CD44+/CD24- CSCs from parental MDA-MB-231 cells in terms of gene expression, tumor cell xenograft formation, and lung metastasis in vitro and in vivo. Single-cell RNA sequencing identified RHBDL2 as a regulator that enhanced spontaneous CD44+/CD24- CSC conversion, whereas knockdown of RHBDL2 expression inhibited YAP/NF-{kappa}B signaling and blocked spontaneous CD44-/CD24- cell conversion to CSCs. These data suggested that the level of CD44-/CD24- tumor cells could predict breast cancer prognosis, metastasis, and response to adjuvant therapy.

physiology↗

bioRxiv2020

U-Net Model for Brain Extraction on Non-human Primates

Brain extraction (a.k.a. skull stripping) is a fundamental step in the neuroimaging pipeline as it can affect the accuracy of downstream preprocess such as image registration, tissue classification, etc. Most brain extraction tools have been designed for and applied to human data and are often challenged by non-human primates (NHP) data. Amongst recent attempts to improve performance on NHP data, deep learning models appear to outperform the traditional tools. However, given the minimal sample size of most NHP studies and notable variations in data quality, the deep learning models are very rarely applied to multi-site samples in NHP imaging. To overcome this challenge, we used a transfer-learning framework that leverages a large human imaging dataset to pretrain a convolutional neural network (i.e. U-Net Model), and then transferred this to NHP data using a small NHP training sample. The resulting transfer-learning model converged faster and achieved more accurate performance than a similar U-Net Model trained exclusively on NHP samples. We improved the generalizability of the model by upgrading the transfer-learned model using additional training datasets from multiple research sites in the Primate Data-Exchange (PRIME-DE) consortium. Our final model outperformed brain extraction routines from popular MRI packages (AFNI, FSL, and FreeSurfer) across a heterogeneous sample from multiple sites in the PRIME-DE with less computational cost (20s~10min). We also demonstrated the transfer-learning process enables the macaque model to be updated for use with scans from chimpanzees, marmosets, and other mammals (e.g. pig). Our model, code, and the skull-stripped mask repository of 136 macaque monkeys are publicly available for unrestricted use by the neuroimaging community at https://github.com/HumanBrainED/NHP-BrainExtraction.

neuroscience↗

bioRxiv2020

Reliability of dynamic network reconfiguration: Impact of code implementation, parameter selection, scan duration, and task condition

Multilayer network models have been proposed as an effective means of capturing the dynamic configuration of distributed neural circuits and quantitatively describing how communities vary over time. Beyond general insights into brain function, a growing number of studies have begun to employ these methods for the study of individual differences. However, test-retest reliabilities for multilayer network measures have yet to be fully quantified or optimized, potentially limiting their utility for individual difference studies. Here, we systematically evaluated the impact of multilayer community detection algorithms, selection of network parameters, scan duration, and task condition on test-retest reliabilities of multilayer network measures (i.e., flexibility, integration, and recruitment). A key finding was that the default method used for community detection by the popular generalized Louvain algorithm can generate erroneous results. Although available, an updated algorithm addressing this issue is yet to be broadly adopted in the neuroimaging literature. Beyond the algorithm, the present work identified parameter selection as a key determinant of test-retest reliability; however, optimization of these parameters and expected reliabilities appeared to be dataset-specific. Once parameters were optimized, consistent with findings from the static functional connectivity literature, scan duration was a much stronger determinant of reliability than scan condition. When the parameters were optimized and scan duration was sufficient, both passive (i.e., resting state, Inscapes, and movie) and active (i.e., flanker) tasks were reliable, although reliability in the movie watching condition was significantly higher than in the other three tasks. The minimal data requirement for achieving reliable measures for the movie watching condition was 20 min, and 30 min for the other three tasks. Our results caution the field against the use of default parameters without optimization based on the specific datasets to be employed - a process likely to be limited for most due to the lack of test-retest samples to enable parameter optimization. HighlightsO_LIDynamic network reliability is highly dependent on many methodological decisions C_LIO_LIThe default multilayer community detection algorithm generates erroneous results C_LIO_LIReliability-optimized intra-/inter-layer coupling parameters are dataset-dependent C_LIO_LIScan duration is a much stronger determinant of reliability than scan condition C_LIO_LIMovies are the most reliable condition, requiring at least 20 min of data C_LI

neuroscience↗

bioRxiv2019

Reliability and Reproducibility of T1-weighted images: comparing MPRAGE and Prospective Motion Correction with Volumetric Navigators

New large neuroimaging studies, such as the Adolescent Brain Cognitive Development study (ABCD) and Human Connectome Project (HCP) Development studies are adopting a new T1-weighted imaging sequence with prospective motion correction (PMC) in favor of the more traditional 3-Dimensional Magnetization-Prepared Rapid Gradient-Echo Imaging (MPRAGE) sequence. Here, we used a developmental dataset (ages 5-21, N=348) from the Healthy Brain Network (HBN) Initiative to directly compare two widely used MRI structural sequences: one based on the Human Connectome Project (MPRAGE) and another based on the ABCD study (MPRAGE+PMC). We aimed to determine if the morphometric measurements obtained from both protocols are equivalent or if one sequence has a clear advantage over the other. The sequences were also compared through quality control measurements. Inter- and intra-sequence reliability were assessed with another set of participants (N=71) from HBN that performed two MPRAGE and two MPRAGE+PMC sequences within the same imaging session, with one MPRAGE (MPRAGE1) and MPRAGE+PMC (MPRAGE+PMC1) pair at the beginning of the session and another pair (MPRAGE2 and MPRAGE+PMC2) at the end of the session. Intraclass correlation coefficients (ICC) scores for morphometric measurements such as volume and cortical thickness showed that intra-sequence reliability is the highest with the two MPRAGE+PMC sequences and lowest with the two MPRAGE sequences. Regarding inter-sequence reliability, ICC scores were higher for the MPRAGE1 - MPRAGE+PMC1 pair at the beginning of the session than the MPRAGE1 - MPRAGE2 pair, possibly due to the higher motion artifacts in the MPRAGE2 run. Results also indicated that the MPRAGE+PMC sequence is robust, but not impervious, to high head motion. For quality control metrics, the traditional MPRAGE yielded better results than MPRAGE+PMC in 5 of the 8 measurements. In conclusion, morphometric measurements evaluated here showed high inter-sequence reliability between the MPRAGE and MPRAGE+PMC sequences, especially in images with low head motion. We suggest that studies targeting hyperkinetic populations use the MPRAGE+PMC sequence, given its robustness to head motion and higher reliability scores. However, neuroimaging researchers studying non-hyperkinetic participants can choose either MPRAGE or MPRAGE+PMC sequences, but should carefully consider the apparent tradeoff between relatively increased reliability, but reduced quality control metrics when using the MPRAGE+PMC sequence.

neuroscience↗

bioRxiv2018

A non-catalytic function of carbonic anhydrase IX contributes to the glycolytic phenotype and pH regulation in human breast cancer cells

The most aggressive and invasive tumor cells often reside in hypoxic microenvironments and rely heavily on rapid anaerobic glycolysis for energy production. This switch from oxidative phosphorylation to glycolysis, along with up-regulation of the glucose transport system, significantly increases the release of lactic acid from cells into the tumor microenvironment. Excess lactate and proton excretion exacerbate extracellular acidification to which cancer cells, but not normal cells, adapt. We have hypothesized that carbonic anhydrases (CAs) play a role in stabilizing both intracellular and extracellular pH to favor cancer progression and metastasis. Here we show that proton efflux (acidification) using the glycolytic rate assay is dependent on both extracellular pH (pHe) and CA IX expression. Yet, isoform selective sulfonamide-based inhibitors of CA IX did not alter proton flux, which suggests that the catalytic activity of CA IX is not necessary for this regulation. Other investigators have suggested the CA IX cooperates with the MCT transport family to excrete protons. To test this possibility, we examined the expression patterns of selected ion transporters and show that members of this family are differentially expressed within the molecular subtypes of breast cancer. The most aggressive form of breast cancer, triple negative breast cancer (TNBC), appears to coordinately express the monocarboxylate transporter 4 (MCT4) and carbonic anhydrase IX (CA IX). This supports a possible mechanism that utilizes the intramolecular H+ shuttle system in CA IX to facilitate proton efflux through MCT4.

cancer biology↗

bioRxiv2018

Evaluating fMRI-Based Estimation of Eye Movement during Naturalistic Viewing

The collection of eye gaze information during functional magnetic resonance imaging (fMRI) is important for monitoring variations in attention and task compliance, particularly for naturalistic viewing paradigms (e.g., movies). However, the complexity and setup requirements of current in-scanner eye-tracking solutions can preclude many researchers from accessing such information. Predictive eye estimation regression (PEER) is a previously developed support vector regression-based method for retrospectively estimating eye gaze from the fMRI signal in the eyes orbit using a 1.5-minute calibration scan. Here, we provide confirmatory validation of the PEER methods ability to infer eye gaze on a TR-by-TR basis during movie viewing, using simultaneously acquired eye tracking data in five individuals (median angular deviation < 2{degrees}). Then, we examine variations in the predictive validity of PEER models across individuals in a subset of data (n=448) from the Child Mind Institute Healthy Brain Network Biobank, identifying head motion as a primary determinant. Finally, we accurately classify which of two movies is being watched based on the predicted eye gaze patterns (area under the curve = .90 {+/-} .02) and map the neural correlates of eye movements derived from PEER. PEER is a freely available and easy-to-use tool for determining eye fixations during naturalistic viewing.

neuroscience↗