Biology subjects

Bates, K.

Publications and source records attributed to Bates, K..

9 recordsLinked to original sources

bioRxiv2026

General prediction of T cell receptor antigen specificity from sequence using AlphaFold 3

The Major Histocompatibility Complex (MHC):peptide:T Cell Receptor (TCR) complex is the most diverse trimolecular interface known in nature and a key trigger for adaptive immunity. TCRs are now routinely sequenced at scale, however decoding their specificity for antigens remains a bottleneck. While in silico approaches have advanced considerably, none enables prediction against "unseen" epitopes, limiting their applicability to a tiny fraction of cases. Here, we show that AlphaFold 3 (AF3) predicts the structures of MHC:peptide:TCR triads with unprecedented accuracy. By applying AF3 to >9,000 TCRs mapped to >1,000 distinct epitopes restricted by >70 MHC class I / II alleles, we identify features that distinguish cognate triads from controls. The resulting model achieves median AUCs of 0.81-0.92 on validation triads unseen by AF3 or during feature selection. Our results reveal that generalized prediction of TCR specificity from sequence is possible, with the potential to greatly accelerate the decoding of immune responses.

immunology↗

bioRxiv2025

Fetoplacental circadian rhythms develop and then synchronize to the mother in utero

Circadian rhythms in gene expression and hormones are ubiquitous across species and differentiated cell types, yet their developmental origins remain poorly understood. This study aimed to determine if daily rhythms can be detected in utero and if they synchronize to the mother. We developed methods to longitudinally monitor PERIOD2 (PER2), a core circadian clock protein, from embryonic day (E)8.5 to E17.5 by restricting PER2::LUCIFERASE (PER2::LUC) expression to the mouse fetoplacental unit (fetus and fetal-derived tissues). In utero fetoplacental bioluminescence imaging showed that PER2 levels rose exponentially during pregnancy, with variable daily peak times that stabilized to dusk by E15.5. Interestingly, pregnancies that did not exhibit daily in utero PER2 variation were more likely to fail. Because maternal glucocorticoids have been implicated in fetal development and synchronizing other circadian tissues, we tested its capability to shift fetoplacental PER2 rhythms in utero. Daily subcutaneous glucocorticoid injections over five days of late pregnancy phase-dependently shifted the fetoplacental PER2 rhythms in utero. Blocking glucocorticoid signaling in vitro reduced synchrony between maternal and fetal placenta by [~]40%. We conclude that in utero daily rhythms gradually develop and synchronize with the mother prior to birth, potentially through glucocorticoid signaling.

developmental biology↗

bioRxiv2025

Learned Geometry, Predicted Binding: Structurally-BasedPrediction of Peptide:MHC Binding Using AlphaFold 3 EnablesCD4 T Cell Epitope Prediction

The accurate prediction of T cell epitope peptides within proteins of interest has a wide range of applications, but is complicated by the multiple determinants of antigenicity, the polymorphism of the Major Histocompatibility Complex (MHC) locus, and the great diversity of possible peptide antigens. Leading in silico methods use a variety of statistical approaches to learn from sequences identified through both in vitro MHC binding and peptide elution studies, but their performance remains imperfect, particularly for MHC II-restricted responses. Here we present MHCIIFold-GNN, an entirely orthogonal solution to this problem that combines three new elements: (i) a highly-multiplexed peptide:MHCII binding assay, (ii) generalizable structural modeling using AlphaFold3, and (iii) transfer learning with a Graph-based Neural Network. Trained exclusively on newly-generated in vitro binders, we show that MHCIIFold-GNN enables state-of-the-art prediction of CD4 T cell epitopes presented by diverse MHC II proteins, on par with non-structural methods that rely on much larger datasets including naturally-processed ligands. Moreover, when MHCIIFold-GNN and a leading non-structural method are combined, we observe unparalleled performance on a held-out test set (11 % boost), underscoring the orthogonality of the methods. These results highlight the power of a new class of structure-informed approaches to the CD4 T cell epitope prediction problem.

bioinformatics↗

bioRxiv2024

The Splice Index as a prognostic biomarker of strength and function in myotonic dystrophy type 1

Myotonic dystrophy type 1 (DM1) is a slowly progressive, multisystemic disorder caused by a CTG repeat expansion in the DMPK 3UTR that leads to global dysregulation of alternative splicing. Here, we employed a composite RNA splicing biomarker called the Myotonic Dystrophy Splice Index (SI), which incorporates 22 disease-specific splice events that sensitively and robustly assesses transcriptomic dysregulation across the disease spectrum. Targeted RNA sequencing was used to derive the SI in 95 muscle biopsies of the tibialis anterior collected from DM1 individuals with baseline (n = 52) and 3-months (n = 37) outcomes. The SI had significant associations with timepoint matched measures of muscle strength and ambulation, including ankle dorsiflexion strength (ADF) and 10-meter run/fast walk speed (Pearson r = -0.719 and -0.680, respectively). Linear regression modeling showed that the combination of baseline ADF and SI was predictive of strength at 3-months (adjusted R2 = 0.830) in our cohort. These results indicate the SI can reliably capture the association of disease-specific RNA mis-splicing to physical strength and mobility and may be predictive of future function.

molecular biology↗

bioRxiv2024

RNA mis-splicing in children with myotonic dystrophy is associated with physical function

ObjectivesDysregulated RNA alternative splicing is the hallmark of myotonic dystrophy type 1 (DM1). However, the association between RNA mis-splicing and physical function in children with the most severe form of disease, congenital myotonic dystrophy (CDM), is unknown. Methods82 participants (42 DM1 adults & 40 CDM children) with muscle biopsies and measures of myotonia, motor function, and strength were combined from five observational studies. Data were normalized and correlated with an aggregate measure of alternative splicing dysregulation, [MBNL]inferred in skeletal muscle biopsies. Multiple linear regression analysis was performed to predict [MBNL]inferred using clinical outcome measures alone. Similar analyses were performed to predict 12-month physical function using baseline metrics. ResultsMyotonia (measured via vHOT) was significantly correlated with RNA mis-splicing in our cross-sectional population of all DM1 individuals; CDM participants alone displayed no myotonia despite a similar range of RNA mis-splicing. Measures of motor performance and muscle strength were significantly associated with [MBNL]inferred in our cohort of all DM1 individuals and when assessing CDM children independently. Multiple linear regression analyses yielded two models capable of predicting [MBNL]inferred from select clinical outcome assessments alone in all subjects (adjusted R2 = 0.6723) or exclusively in CDM children (adjusted R2 = 0.5875). InterpretationOur findings establish significant correlations between skeletal muscle performance and a composite measure of alternative splicing dysregulation, [MBNL]inferred, in DM1. The strength of these correlations and the development of the predictive models will assist in designing efficacious clinical trials for individuals with DM1, particularly CDM.

genetics↗

bioRxiv2023

Fetal daily rhythms develop during pregnancy and entrain to the maternal circadian system

Circadian rhythms in gene expression and hormones are ubiquitous across species and differentiated cell types. This study aimed to determine when daily rhythms begin in the fetus and synchronize to the mother. We developed methods to monitor the expression of fetal PERIOD2 (PER2), a core circadian clock protein, in mice longitudinally from embryonic day (E)8.5 to E17.5 through in utero bioluminescence imaging. We found that embryonic PER2 expression increased rapidly throughout pregnancy and exhibited day-night rhythms from the start of our recordings at E8.5. The daily peak time of PER2 varied between pregnancies until it reliably peaked at night and synchronized to the mother starting around E15.5. Loss of fetal circadian rhythms associated with pregnancies that ultimately failed. Because maternal glucocorticoids have been implicated in fetal development and synchronizing circadian tissues, we tested their sufficiency to shift fetal daily rhythms in utero. Daily glucocorticoids injections over five days of late pregnancy advanced fetal PER2 rhythms in utero and blocking glucocorticoid signaling in vitro reduced PER2 synchrony between the maternal and fetal placenta by [~]40%. We conclude that fetal daily rhythms arise early in pregnancy and then synchronize with the maternal rhythm prior to birth depending, in part, on glucocorticoid signaling.

developmental biology↗

bioRxiv2021

Deep learning for robust and flexible tracking in behavioral studies for C. elegans

Robust and accurate behavioral tracking is essential for ethological studies. Common methods for tracking and extracting behavior rely on user adjusted heuristics that can significantly vary across different individuals, environments, and experimental conditions. As a result, they are difficult to implement in large-scale behavioral studies with complex, heterogenous environmental conditions. Recently developed deep-learning methods for object recognition such as Faster R-CNN have advantages in their speed, accuracy, and robustness. Here, we show that Faster R-CNN can be employed for identification and detection of Caenorhabditis elegans in a variety of life stages in complex environments. We applied the algorithm to track animal speeds during development, fecundity rates and spatial distribution in reproductive adults, and behavioral decline in aging populations. By doing so, we demonstrate the flexibility, speed, and scalability of Faster R-CNN across a variety of experimental conditions, illustrating its generalized use for future large-scale behavioral studies.

bioengineering↗

bioRxiv2018

Fast, versatile, and quantitative annotation of complex images

We report a generic smartphone app for quantitative annotation of complex images. The app is simple enough to be used by children, and annotation tasks are distributed across app users, contributing to efficient annotation. We demonstrate its flexibility and speed by annotating >30,000 images, including features of rice root growth and structure, stem cell aggregate morphology, and complex worm (C. elegans) postures, for which we show that the speed of annotation is >130-fold faster than state-of-the-art techniques with similar accuracy.

bioengineering↗

bioRxiv2018

Development and worldwide use of a non-lethal and minimal population-level impact protocols for the isolation of chytrids from amphibians

Parasitic chytrid fungi have emerged as a significant threat to amphibian species worldwide, necessitating the development of techniques to isolate these pathogens into sterile culture for research purposes. However, early methods of isolating chytrids from their hosts relied on killing amphibians. We modified a pre-existing protocol for isolating chytrids from infected animals to use toe clips and biopsies from toe webbing rather than euthanizing hosts, and distributed the protocol to interested researchers worldwide as part of the BiodivERsA project RACE - here called the RML protocol. In tandem, we developed a lethal procedure for isolating chytrids from tadpole mouthparts. Reviewing a database of use a decade after their inception, we find that these methods have been widely applied across at least 5 continents, 23 countries and in 62 amphibian species, and have been successfully used to isolate chytrids in remote field locations. Isolation of chytrids by the non-lethal RML protocol occured in 18% of attempts with 207 fungal isolates and three species of chytrid being recovered. Isolation of chytrids from tadpoles occured in 43% of attempts with 334 fungal isolates of one species (Batrachochytrium dendrobatidis) being recovered. Together, these methods have resulted in a significant reduction and refinement of our use of threatened amphibian species and have improved our ability to work with this important group of emerging fungal pathogens.

ecology↗