Biology subjects

Hartman, E.

Publications and source records attributed to Hartman, E..

12 recordsLinked to original sources

bioRxiv2026

Learning proteomic disease trajectories with flow matching

High-throughput proteomics has enabled detailed characterization of molecular states across health and disease. However, biological systems are inherently dynamic and methods for reconstructing continuous proteome changes remain limited. Here, we introduce proteome velocity, a framework for inferring continuous proteome trajectories from cross-sectional or sparsely sampled proteomics data using flow matching, in which a neural network learns velocity fields over proteome space. Proteome velocity estimates how rapidly and in which direction protein abundances change along a biological progression, such as disease. In mouse sepsis, covariate-conditioned velocity models resolved tissue- and pathogen-specific proteome trajectories and identified inflammatory proteins with distinct temporal activation patterns across infection routes and organ systems. In clinical COVID-19 plasma proteomes, inferred trajectories separated into distinct velocity programs associated with disease severity. These results show how generative trajectory models can transform cross-sectional proteomics data into interpretable, protein-resolved representations of molecular progression.

bioinformatics↗

bioRxiv2025

Mining the endogenous peptidome for peptide binders with deep learning-driven optimization and molecular simulations

Peptides are short amino-acid chains that mediate essential biological processes, including antimicrobial defence, immune modulation and cell signalling. Their high degree of modularity, biocompatibility and capacity to bind proteins with high specificity make them attractive therapeutic candidates. However, identifying peptides that bind and modulate the function of specific proteins remains challenging due to the immense size of the peptide sequence space. To adress this challenge, we developed BoPep (Bayesian Optimization for Peptides), an end-to-end modular framework that effectively navigates the landscape of peptide-protein interactions by directing the search toward informative regions of sequence space and prioritizes candidates with high binding potential. By focusing computational effort where it is most informative and using calibrated uncertainty to balance exploration and exploitation, BoPep reduces the number of expensive docking evaluations by orders of magnitudes. We demonstrate the utility of BoPep by applying it to three sources of peptides: endogenous proteolytic fragments from clinical wound fluids, the complete human proteome, and a de novo design peptide landscape generated by diffusion-based backbone sampling. Using these sources, we uncover novel encrypted peptide classes that bind CD14 and identify peptides that neutralize the hemolytic activity of pneumolysin, a major bacterial virulence factor. Together, these findings show that BoPep accelerates the identification of testable therapeutic leads from large and diverse peptide collections. BoPep is available at GitHub.

bioinformatics↗

bioRxiv2023

Community-based peptidomic clustering reveals pathogen-specific proteolytic signatures in wounds

Recent advances in mass spectrometry-based peptidomics has catalyzed the identification and quantification of thousands endogenous peptides across diverse biological systems. However, the large theoretical peptidomic landscape and high proportion of missing values poses several challenges for downstream analyses and limits the comparability of clinical samples. Here, we present a generalizable computational workflow that clusters peptides with overlapping sequences to reduce the dimensionality of peptidomic data, improve the definition of protease cut-sites, enhance inter-sample comparability, and enable the implementation of reliable and large-scale data analysis methods akin to those employed in other omics fields. We showcase the algorithm by performing large-scale quantitative analysis of wound fluid peptidomes of highly defined porcine wounds and human clinical non-healing wounds. The analysis revealed signature phenotype-specific peptide regions reflecting pathogen-specific proteolytic activity at the earliest stages of colonization, resulting in novel class of potential peptide cluster-based biomarkers.

bioinformatics↗

bioRxiv2023

Explainable machine learning for the identification of proteome states via the data processing kitchen sink

The application of machine learning algorithms to facilitate the understanding of changes in proteome states has emerged as a promising methodology in proteomics research. Unfortunately, these methods can prove difficult to interpret, as it may not be immediately obvious how models reach their predictions. We present the data processing kitchen sink (DPKS) which provides reproducible access to classic statistical methods and advanced explainable machine learning algorithms to build highly accurate and fully interpretable predictive models. In DPKS, explainable machine learning methods are used to calculate the importance of each protein towards the prediction of a model for a particular proteome state. The calculated importance of each protein can enable the identification of proteins that drive phenotypic change in a data-driven manner while classic techniques rely on arbitrary cutoffs that may exclude important features from consideration. DPKS is a free and open source Python package available at https://github.com/InfectionMedicineProteomics/DPKS.

bioinformatics↗

bioRxiv2023

Separating cognitive and motor processes in the behaving mouse

The cognitive processes supporting complex animal behavior are closely associated with ubiquitous movements responsible for our posture, facial expressions, ability to actively sample our sensory environments, and other critical processes. These movements are strongly related to neural activity across much of the brain and are often highly correlated with ongoing cognitive processes, making it challenging to dissociate the neural dynamics that support cognitive processes from those supporting related movements. In such cases, a critical issue is whether cognitive processes are separable from related movements, or if they are driven by common neural mechanisms. Here, we demonstrate how the separability of cognitive and motor processes can be assessed, and, when separable, how the neural dynamics associated with each component can be isolated. We establish a novel two-context behavioral task in mice that involves multiple cognitive processes and show that commonly observed dynamics taken to support cognitive processes are strongly contaminated by movements. When cognitive and motor components are isolated using a novel approach for subspace decomposition, we find that they exhibit distinct dynamical trajectories. Further, properly accounting for movement revealed that largely separate populations of cells encode cognitive and motor variables, in contrast to the mixed selectivity often reported. Accurately isolating the dynamics associated with particular cognitive and motor processes will be essential for developing conceptual and computational models of neural circuit function and evaluating the function of the cell types of which neural circuits are composed.

neuroscience↗

bioRxiv2023

Interpreting biologically informed neural networks for enhanced biomarker discovery and pathway analysis

The advent of novel methods in mass spectrometry-based proteomics allows for the identification of biomarkers and biological pathways which are crucial for the understanding of complex diseases. However, contemporary analytical methods often omit essential information, such as protein abundance and protein co-regulation, and therefore miss crucial relationships in the data. Here, we introduce a generalized workflow that incorporates proteins, their abundances, and associated pathways into a deep learning-based methodology to improve biomarker identification and pathway analysis through the creation and interpretation of biologically informed neural networks (BINNs). We successfully employ BINNs to differentiate between two subphenotypes of septic acute kidney injury (AKI) and COVID-19 from the plasma proteome and utilize feature attribution-methods to introspect the networks to identify which proteins and pathways are important for distinguishing between subphenotypes. Compared to existing methods, BINNs achieved the highest predictive accuracy and revealed that metabolic processes were key to differentiating between septic AKI subphenotypes, while the immune system was more important to the classification of COVID-19 subphenotypes. The methodology behind creating, interpreting, and visualizing BINNs were implemented in a free and open source Python-package: https://github.com/InfectionMedicineProteomics/BINN.

bioinformatics↗

bioRxiv2023

Protein aggregation in wound fluid confines bacterial lipopolysaccharide and reduces inflammation

Bacterial lipopolysaccharide (LPS) induces the rapid formation of protein aggregates in human wound fluid. We aimed to define such LPS-induced aggregates and the functional consequences of protein aggregation using a combination of mass spectrometry analyses, biochemical imaging, and experimental animal models. We show that such wound-fluid aggregates contain a multitude of protein classes, including sequences from coagulation factors, annexins, histones, antimicrobial proteins/peptides, and apolipoproteins. Proteins and peptides with a high aggregation propensity were identified, and selected components were verified biochemically by western blot analysis. Staining by thioflavin T and the Amytracker probe demonstrated the presence of amyloid-like aggregates formed after exposure to LPS in vitro in human wound fluid and in vivo in porcine wound models. Using NF-{kappa}B-reporter mice and IVIS bioimaging, we show that such wound-fluid LPS aggregates induce a significant reduction in local inflammation compared with LPS in plasma. The results show that protein/peptide aggregation is a mechanism for confining LPS and reducing inflammation and further underscore the connection between host defense and amyloidogenesis.

immunology↗

bioRxiv2022

Systematic Engineering of Virus-Like Particles to Identify Self-Assembly Rules for Shifting Particle Size

Virus-like particles (VLPs) are promising scaffolds for biomaterials as well as diagnostic and therapeutic applications. However, there are some key challenges to be solved, such as the ability to engineer alternate sizes for varied use cases. To this end, we created a library of MS2 VLP variants at two key residues in the coat protein which have been implicated as important to controlling VLP size and geometry. By adapting a method for systematic mutagenesis coupled with size-based selections and high-throughput sequencing as a readout, we developed a quantitative assessment of two residues in MS2 coat protein that govern the size shift in MS2 VLPs. We then applied the strategy to the equivalent residues in Q{beta} VLPs, an MS2 homolog, and demonstrate that the analogous pair of residues are also able to impact VLP size and shape. These results underscore the power of fitness landscapes in identifying critical features for assembly.

bioengineering↗

bioRxiv2022

Within-species control reveals novel immune response genes in Caenorhabditis elegans

The nematode Caenorhabditis elegans is a simple model host for studying the interaction between bacterial pathogens and the metazoan innate immune system. In the last two decades, much focus has been given to studying intestinal infection in the worm by the clinical strain Pseudomonas aeruginosa PA14. Powerful genetic and molecular tools in both species facilitate the identification and analysis of bacterial virulence factors as well as host defense factors. However, findings from these studies are confounded by the use of the genetically, metabolically, and physiologically divergent E. coli OP50 as a food source and as the non-virulent control. Here we report the use of P. aeruginosa z11 strain as a preferable control for PA14 infection studies. We demonstrate that many aspects of worm behavior, health-span, longevity, food attraction, brood size, lifespan, and pathogen avoidance and survival are not affected in z11. We show that the use of z11 as a control for transcriptomics analysis can increase the discovery power. In particular, we identified three novel pathogenesis-related genes in C. elegans. Our findings accentuate the importance of choosing an appropriate control environment for host-pathogen studies.

genomics↗

bioRxiv2022

Limb Specific Failure of Proliferation and Translation in the Mesenchyme Leads to Skeletal Defects in Diamond Blackfan Anemia

AbstractRibosomopathies are a class of disorders caused by defects in the structure or function of the ribosome and characterized by tissue-specific abnormalities. Diamond Blackfan anemia (DBA) arises from different mutations, predominantly in genes encoding ribosomal proteins (RPs). Apart from the anemia, skeletal defects are among the most common anomalies observed in patients with DBA, but they are virtually restricted to radial ray and other upper limb defects. What leads to these site-specific skeletal defects in DBA remains a mystery. Using a novel mouse model for RP haploinsufficiency, we observed specific, differential defects of the limbs. Using complementary in vitro and in vivo approaches, we demonstrate that reduced WNT signaling and subsequent increased {beta}-catenin degradation in concert with increased expression of p53 contribute to mesenchymal lineage failure. We observed differential defects in the proliferation and differentiation of mesenchymal stem cells (MSCs) from the forelimb versus the hind limbs of the RP haploinsufficient mice that persisted after birth and were partially rescued by allelic reduction of Trp53. These defects are associated with a global decrease in protein translation in RP haploinsufficient MSCs, with the effect more pronounced in cells isolated from the forelimbs. Together these results demonstrate translational differences inherent to the MSC, explaining the site-specific skeletal defects observed in DBA.

cell biology↗

bioRxiv2021

Peptimetric: Quantifying and visualizing differences in peptidomic data

Finding new sustainable means of diagnosing and treating diseases is one of the most pressing issues of our time. In recent years, several endogenous peptides have been found to be both excellent biomarkers for many diseases and to possess important physiological roles which may be utilized in treatments. The detection of peptides has been facilitated by the rapid development of biological mass spectrometry and now the combination of fast and sensitive high resolution MS instruments and stable nano HP-LC equipment sequences thousands of peptides in one single experiment. In most research conducted with these advanced systems, proteolytically cleaved proteins are analyzed and the specific peptides are identified by software dedicated for protein quantification using different proteomics workflows. Analysis of endogenous peptides with peptidomics workflows also benefit from the novel sensitive and advanced instrumentation, however, the generated peptidomic data is vast and subsequently laborious to visualize and examine, creating a bottleneck in the analysis. Therefore, we have created Peptimetric, an application designed to allow researchers to investigate and discover differences between peptidomic samples. Peptimetric allows the user to dynamically and interactively investigate the proteins, peptides, and some general characteristics of multiple samples, and is available as a web application at https://peptimetric.herokuapp.com. To illustrate the utility of Peptimetric, weve applied it to a peptidomic dataset of 15 urine samples from diabetic patients and corresponding data from healthy subjects.

bioinformatics↗

bioRxiv2019

Coxiella burnetii epitope-specific T-cell responses in chronic Q fever patients

Infection with Coxiella burnetii, the causative agent of Q fever, can result in life-threatening persistent infection. Reactogenicity hinders worldwide implementation of the only licensed human Q fever vaccine. We previously demonstrated long-lived immunoreactivity in individuals with past symptomatic and asymptomatic Coxiella infection (convalescents) to promiscuous HLA-class II C. burnetii epitopes, providing the basis for a novel T-cell-targeted subunit vaccine. Here we investigated in a cohort of 22 individuals with persistent infection (chronic Q fever) whether they recognize the same set of epitopes, or distinct epitopes that could be candidates for a therapeutic vaccine or aid in the diagnosis of persistent infection.\n\nIndividuals with chronic Q fever showed strong class II epitope-specific cultured ELISpot responses largely overlapping with the peptide repertoire identified previously for convalescents. Five additional peptides were recognized more frequently by chronic subjects, but there was no combination of epitopes uniquely recognized by or non-reactive in chronic Q fever subjects. Consistent with more recent/prolonged exposure, we found, however, stronger direct ex vivo responses to whole-cell C. burnetii and individual peptides in direct ELISpot than in convalescents.\n\nIn conclusion, we have validated and expanded a previously published set candidate epitopes for a novel T-cell targeted subunit Q fever vaccine in the context of chronic Q fever patients and demonstrated that they successfully mounted a T-cell response comparable to that of convalescents. Finally, we demonstrate that individuals treated for chronic Q fever mount a broader ex vivo response to class II epitopes than convalescents, which could be explored for diagnostic purposes.

immunology↗