Biology subjects

Kaufman, J.

Publications and source records attributed to Kaufman, J..

At least 19 recordsLinked to original sources

bioRxiv2026

Simplifying principles that underlie the highly complex peptide motif of the promiscuous chicken class I molecule, BF2*21:01

In chickens and humans, classical class I molecules of the major histocompatibility complex (MHC) have a spectrum of correlated properties, including cell surface expression and peptide repertoire. Understanding the impact of these promiscuous generalists and fastidious specialists is of considerable interest. The promiscuous BF2 molecule from the chicken B21 haplotype, BF2*21:01, binds a wide range of peptides by remodelling the peptide-binding site, allowing co-variation of the anchor residues at peptide positions P2 and Pc-2, and binding of an anchor residue at Pc. Using in vitro refolding assays with peptides and peptide libraries, determining thermostability and crystal structures, and analysing a chicken B21 cell line by immunopeptidomics, we found that BF2*21:01 will accommodate many possible combinations at P2 and Pc-2, as well as several hydrophobic amino acids at Pc. However, marked preferences for particular peptide lengths, particular amino acids at the three anchor residues, and particular combinations of amino acids at P2 and Pc-2 as well as at P3 and Pc-3 together lead to high frequencies of major peptides while still allowing the possibility of presenting a wide peptide repertoire. These simplifying principles may eventually allow predictions of pathogen peptides with stable binding for this iconic promiscuous class I molecule, as well as providing the data for more sophisticated peptide prediction methods.

immunology↗

bioRxiv2025

MHC-I diversity enables rapid adaptation during viral pandemic in wild rabbit populations

Emerging diseases can have devastating consequences for wild species, with long-term effects depending on the ability of the host to evolve resistance. Here, we show that major histocompatibility complex (MHC) genes provided standing genetic variation that enabled rabbits to mount a rapid evolutionary response to the myxoma virus pandemic that began in the 1950s. Using historical and modern specimens starting in 1865 and spanning the pandemic, we found strong parallel shifts in MHC-I allele frequencies across Australia, Britain and France, alongside population-specific signals. These evolutionary shifts appear to have altered antigen presentation to T cells. Our results provide evidence that MHC-I is under strong selection in natural populations during a pandemic, and that the high polymorphism of MHC may have contributed to the evolutionary rescue of these populations.

evolutionary biology↗

bioRxiv2025

A major histocompatibility complex (MHC) class II molecule that binds the same viral pathogen peptide with both nonamer and decamer core sequences for presentation to T cells

Classical molecules encoded by the major histocompatibility complex (MHC) are central to immune responses. Compared to typical mammals, the chicken MHC is small and simple, determining life or death from economically-important pathogens like avian influenza virus and Mareks disease virus (MDV). Several genes within the tightly-linked chicken MHC have been suggested to determine resistance and susceptibility to MDV, but it was a surprise to find that the dominantly-expressed class II molecule from the resistant B2 haplotype employed a novel peptide-binding mode with a decamer core sequence compared to the susceptible B19 haplotype with a typical nonamer core. We examined the crystal structure of the dominantly-expressed class II molecule from another resistant haplotype, B21, which is extremely frequent in commercial chicken flocks, to find that it bound the same MDV peptide with both nonamer and decamer cores, revealing an unexpected plasticity of binding that potentially increases the immune response to this devastating pathogen.

immunology↗

bioRxiv2024

A single-nucleus census of immune and non-immune cell types for the major immune organ systems of chicken.

In the avian host, comprehensively cataloging immune cell types, their transcriptome profiles, and varying molecular responses to pathogen challenges are necessary steps toward a better understanding of the interplay between genetics and disease resilience. We present a first nuclei atlas of immune cell types derived from the three main immune organs of layer chickens, including spleen, bursa, and thymus. In bursa we also present, an accounting of cell type activation with the bacterial toxin lipopolysaccharide (LPS). Our analysis includes 36,370 total nuclei and 16, 12, and 12 transcriptionally distinct clusters for spleen, bursa, and thymus, respectively. We discover nuclei molecular profiles that uniquely distinguish states of the transcriptome within cell type that could serve as new means to characterize avian immune subtypes. We further subcluster refined immune cell type classifications, specifically highlighting the transcriptomic diversity of B and T cell subtypes. In the bursa, inferred intercellular communication and signaling pathway enrichment analyses across immune and non-immune cell types demonstrate the unappreciated complexity of the B cell repertoire in a model mimicking systemic bacterial infection. This census of all cell types in both primary and one major secondary avian immune organ system, although preliminary, provides a first review of how nuclei transcribe numerous genes, known and unknown, a critical prerequisite for the study avian immunogenetics by cell type.

immunology↗

bioRxiv2024

Measles virus co-opts epithelial-to-mesenchymal transition and live cell extrusion to exit human airway epithelia

Measles virus (MeV) is a highly contagious respiratory virus transmitted via aerosols. To understand how MeV exits the airways of an infected host, we use unpassaged primary cultures of human airway epithelial cells (HAE). MeV typically remains cell-associated in HAE and forms foci of infection, termed infectious centers, by directly spreading cell-to-cell. We previously described the phenomenon in which infectious centers detach en masse from HAE and remain viable. Here we investigate the mechanism of this cellular detachment. Via immunostaining, we observed loss of tight junction and cell adhesion proteins within infectious centers. These morphological changes indicate activation of epithelial-to-mesenchymal transition (EMT). EMT can contribute to wound healing in respiratory epithelia by mobilizing nearby cells. Inhibiting TGF-{beta}, and thus EMT, reduced infectious center detachment. Compared to uninfected cells, MeV-infected cells also expressed increased levels of sphingosine kinase 1 (SK1), a regulator of live cell extrusion. Live cell extrusion encourages cells to detach from respiratory epithelia by contracting the actomyosin of neighboring cells. Inhibition or induction of live cell extrusion impacted infectious center detachment rates. Thus, these two related pathways contributed to infectious center detachment in HAE. Detached infectious centers contained high titers of virus that may be protected from the environment, allowing the virus to live on surfaces longer and infect more hosts. This mechanism may contribute to the highly contagious nature of MeV. IMPORTANCEMeasles virus (MeV) is an extremely contagious respiratory pathogen that continues to cause large, disruptive outbreaks each year. Here, we examine a phenomenon that may help explain the contagious nature of MeV: detachment of MeV-infected cells. MeV spreads cell-to-cell in human airway epithelial cells (HAE) to form groups of infected cells, termed "infectious centers". We reported that infectious centers ultimately detach from HAE as a unit, carrying high titers of virus. Viral particles within cells may be more protected from environmental conditions, such as ultraviolet radiation and desiccation. We identified two host pathways, epithelial-to-mesenchymal transition and live cell extrusion, that contribute to infectious center detachment. Perturbing these pathways altered the kinetics of infectious centers detachment. These pathways influence one another and contribute to epithelial wound healing, suggesting infectious center detachment may be a usurped consequence of the hosts response to infection that benefits MeV by increasing its transmissibility between hosts.

microbiology↗

bioRxiv2024

Human Breast Milk EVs Mitigate Endothelial Dysfunction: Preliminary Study

BackgroundEndothelial cell (EC) dysfunction is an early indicator of failing vascular integrity, leading to various cardiovascular diseases. Toll-like receptor 4 (TLR4) activation is a key mechanism. Milk-derived extracellular vesicles (EVs) are known for their anti-inflammatory properties, particularly in suppressing TLR4 activation in damaged intestinal epithelial cells. This study explores the therapeutic potential of human breast milk EVs (bEVs) in EC dysfunction related to cardiovascular diseases. MethodsHuman breast milk EVs (bEVs) were isolated from healthy nursing mothers using ultracentrifugation. bEVs were applied to LPS-treated HUVECs, and the expression of inflammatory markers was measured using qPCR and western blotting. Angiogenesis was assessed via a wound assay. Additionally, bEVs were orally administered weekly for six weeks to high-fat diet-induced obese mice and lean mice. Metabolic phenotype characteristics and EC-dependent vasorelaxation were evaluated. ResultsbEV pre-treatment inhibited LPS-induced expression of inflammatory genes, including IL-6, IL-1b and VCAM-1. It also suppressed phospho-NFkB and TLR4 protein expression. bEVs enhanced EC migration, significantly increasing wound closure. Oral administration of bEVs restored impaired EC-dependent vasorelaxation in the mesenteric artery of obese mice, though metabolic parameters remained unchanged. ConclusionOur findings demonstrate the beneficial effects of bEVs on EC dysfunction, highlighting their potential as novel therapeutics for cardiovascular diseases. Future studies will focus on identifying specific bEV cargos and further evaluating their therapeutic effects on EC dysfunction.

cell biology↗

bioRxiv2022

A Community Challenge to Predict Clinical Outcomes After Immune Checkpoint Blockade in Non-Small Cell Lung Cancer

PurposePredictive biomarkers of immune checkpoint inhibitors (ICIs) efficacy are currently lacking for non-small cell lung cancer (NSCLC). Here, we describe the results from the Anti-PD-1 Response Prediction DREAM Challenge, a crowdsourced initiative that enabled the assessment of predictive models by using data from two randomized controlled clinical trials (RCTs) of ICIs in first-line metastatic NSCLC. MethodsParticipants developed and trained models using public resources. These were evaluated with data from the CheckMate 026 trial (NCT02041533), according to the model-to-data paradigm to maintain patient confidentiality. The generalizability of the models with the best predictive performance was assessed using data from the CheckMate 227 trial (NCT02477826). Both trials were phase III RCTs with a chemotherapy control arm, which supported the differentiation between predictive and prognostic models. Isolated model containers were evaluated using a bespoke strategy that considered the challenges of handling transcriptome data from clinical trials. ResultsA total of 59 teams participated, with 417 models submitted. Multiple predictive models, as opposed to a prognostic model, were generated for predicting overall survival, progression-free survival, and progressive disease status with ICIs. Variables within the models submitted by participants included tumor mutational burden (TMB), programmed death ligand 1 (PD-L1) expression, and gene-expression-based signatures. The bestperforming models showed improved predictive power over reference variables, including TMB or PD-L1. ConclusionThis DREAM Challenge is the first successful attempt to use protected phase III clinical data for a crowdsourced effort towards generating predictive models for ICIs clinical outcomes and could serve as a blueprint for similar efforts in other tumor types and disease states, setting a benchmark for future studies aiming to identify biomarkers predictive of ICIs efficacy. Context summaryO_ST_ABSKey objectiveC_ST_ABSNot all patients with non-small cell lung cancer (NSCLC) eligible for immune checkpoint inhibitor (ICIs) respond to treatment, but accurate predictive biomarkers of ICIs clinical outcomes are currently lacking. This crowdsourced initiative enabled the robust assessment of predictive models using data from two randomized clinical trials of first-line ICI in metastatic NSCLC. Knowledge generatedModels submitted indicate that a combination of programmed death ligand 1 (PD-L1), tumor mutational burden (TMB), and immune gene signatures might be able to identify patients more likely to respond to ICIs. TMB and PD-L1 seemed important to predict progression-free survival and overall survival. Mechanisms including apoptosis, T-cell crosstalk, and adaptive immune resistance appeared essential to predict response. Relevance

cancer biology↗

bioRxiv2022

Retrotransposons facilitates tissue specific horizontal Transfer of circulating tumor DNA

A variety of organisms have been shown to have altered physiology or developed pathology due to gene transfer, but mammals have never been shown to do so. Here, we show that circulating tumor DNA (ct) can promote cell-specific horizontal gene transfer (HGT) between human cancer cells and explain the mechanisms behind this phenomenon. Once ctDNA enters the host cell, it migrates to the nucleus and integrates into the cells genome, thereby transferring its genetic information. We determine that retrotransposons of the ERVL, SINE, and LINE families are necessary for cell targeting and the integration of ctDNA into host DNA. Using chemically synthesized retrotransposons, we found that AluSp and MER11C reproduced multiple myelomas (MM) ctDNAs cell targeting and integration into MM cells. We also discovered that ctDNA might, as a result of HGT, influence the treatment response of multiple myeloma and pancreatic cancer models. Overall, this is the first study to show that retrotransposon-directed HGT can promote genetic material transfer in cancer. There is, however, a broader impact of our findings than just cancer since cell-free DNA has also been found in physiological and other pathological conditions as well. Furthermore, with the discovery of transposons-mediated tissue-specific targeting, a new avenue for the delivery of genes and therapies will emerge.

cancer biology↗

bioRxiv2022

Minor differences in peptide presentation between chicken MHC class I molecules can explain differences in disease susceptibility

The highly polymorphic classical major histocompatibility complexes (MHCs) can confer resistance or susceptibility to diseases. The chicken MHC is known to confer decisive resistance or susceptibility to various economically-important pathogens, including the iconic oncogenic herpesvirus that causes Mareks disease (MD). Only one classical class I gene, BF2, is expressed at a high level in chickens, so it was relatively easy to discern a hierarchy from well-expressed thermostable fastidious specialist alleles to promiscuous generalist alleles that are less stable and expressed less on the cell surface. The BF2*1901 is more highly expressed and more thermostable than the closely-related BF2*1501, but the data for peptide repertoire available did not obviously correlate as expected. Here, we confirm for newly-developed chicken lines that the chicken haplotype B15 confers resistance to MD compared to B19. Using gas phase sequencing of peptide pools, and using immunopeptidomics involving mass spectroscopy, we find that the BF2*1901 binds a greater variety of amino acids in some anchor positions than BF2*1501. However, by X-ray crystallography, we find that the peptide-binding groove of BF2*1901 is narrower and shallower. Though the self-peptides bound to BF2*1901 may appear more various than those of BF2*1501, the structures show that the wider and deeper peptide-binding groove of BF2*1501 allows it to tightly accept many more peptides overall, correlating with the expected hierarchies for expression level, thermostability and MD resistance. Moreover, the 2 helix of BF2*1901 is higher than BF2*1501, potentially reducing the number of T cell clones that can recognize this fastidious class I molecule. IMPORTANCEDisease susceptibility mechanism is complicated, but chicken infection of Mareks disease (MD) is one of ideal models, considering the only one highly expressed classical MHC class I, BF2. The different susceptibility of the two close alleles BF2*1901 and BF2*1501 with minor difference of expression and thermostability is still unfathomed. Here, we confirm B15 chicken confers resistance to MD compared to B19. But the BF2*1901 binds a broader variety of anchoring peptides than BF2*1501. This mystery was solved by the structural determination of the two molecules with one similar peptide. The wider and deeper peptide-binding groove of BF2*1501 allows it to tightly accept many more peptides overall, which is concordant to the expected hierarchies for expression level, thermostability and MD resistance.

immunology↗

bioRxiv2022

Predicting Epitope Candidates for SARS-CoV-2

Epitopes are short amino acid sequences that define the antigen signature to which an antibody binds. In light of the current pandemic, epitope analysis and prediction is paramount to improving serological testing and developing vaccines. In this paper, we leverage known epitope sequences from SARS-CoV, SARS-CoV-2 and other Coronaviridae and use those known epitopes to identify additional antigen regions in 62k SARS-CoV-2 genomes. Additionally, we present epitope distribution across SARS-CoV-2 genomes, locate the most commonly found epitopes, discuss where epitopes are located on proteins, and how epitopes can be grouped into classes. We also discuss the mutation density of different regions on proteins using a big data approach. We find that there are many conserved epitopes between SARS-CoV-2 and SARS-CoV, with more diverse sequences found in Nucleoprotein and Spike Glycoprotein.

bioinformatics↗

bioRxiv2020

Analysis and Forecasting of Global of RT-PCR Primers for SARS-CoV-2

Rapid tests for active SARS-CoV-2 infections rely on reverse transcription polymerase chain reaction (RT-PCR). RT-PCR uses reverse transcription of RNA into complementary DNA (cDNA) and amplification of specific DNA (primer and probe) targets using polymerase chain reaction (PCR). The technology makes rapid and specific identification of the virus possible based on sequence homology of nucleic acid sequence and is much faster than tissue culture or animal cell models. However the technique can lose sensitivity over time as the virus evolves and the target sequences diverge from the selective primer sequences. Different primer sequences have been adopted in different geographic regions. As we rely on these existing RT-PCR primers to track and manage the spread of the Coronavirus, it is imperative to understand how SARS-CoV-2 mutations, over time and geographically, diverge from existing primers used today. In this study, we analyze the performance of the SARS-CoV-2 primers in use today by measuring the number of mismatches between primer sequence and genome targets over time and spatially. We find that there is a growing number of mismatches, an increase by 2% per month, as well as a high specificity of virus based on geographic location.

bioinformatics↗

bioRxiv2020

Differential host response to a herpesvirus: Mareks disease virus peptides on chicken MHC class II molecules are derived from only a few genes and illustrate a new mode of peptide binding

Viral diseases pose major threats to humans and other animals, including the billions of chickens that are an important food source as well as a public health concern due to zoonotic pathogens. Unlike humans and other typical mammals, the major histocompatibility complex (MHC) of chickens can confer decisive resistance or susceptibility to many viral diseases. An iconic example is Mareks disease, caused by an oncogenic herpesvirus with over 100 genes. Classical MHC class I and class II molecules present antigenic peptides to T lymphocytes, and it has been hard to understand how such MHC molecules could be involved in susceptibility to Mareks disease, given the potential number of peptides from over 100 genes. We used a new in vitro infection system and immunopeptidomics to determine peptide motifs for the two class II molecules expressed by the MHC haplotype B2, which is known to confer resistance to Mareks disease. Surprisingly, we found that the vast majority of viral peptide epitopes presented by chicken class II molecules arise from only four viral genes, nearly all having the peptide motif for BL2, the dominantly-expressed class II molecule in chickens. We expressed BL2 linked to several MDV peptides, and determined one X-ray crystal structure, showing how a single small amino acid in the binding site causes a crinkle in the peptide, leading to core binding peptide of ten amino acids, compared to the nine amino acids in all other reported class II molecules. The limited number of potential T cell epitopes from such a complex virus can explain the differential MHC-determined resistance to MDV, but raises questions of mechanism and opportunities for vaccine targets in this important food species, as well as providing a basis for understanding class II molecules in other species including humans.

immunology↗

bioRxiv2020

Monitoring the microbiome for food safety and quality using deep shotgun sequencing

In this work, we hypothesized that shifts in the food microbiome can be used as an indicator of unexpected contaminants or environmental changes. To test this hypothesis, we sequenced total RNA of 31 high protein powder (HPP) samples of poultry meal pet food ingredients. We developed a microbiome analysis pipeline employing a key eukaryotic matrix filtering step that improved microbe detection specificity to >99.96% during in silico validation. The pipeline identified 119 microbial genera per HPP sample on average with 65 genera present in all samples. The most abundant of these were Bacteroides, Clostridium, Lactococcus, Aeromonas, and Citrobacter. We also observed shifts in the microbial community corresponding to ingredient composition differences. When comparing culture-based results for Salmonella with total RNA sequencing, we found that Salmonella growth did not correlate with multiple sequence analyses. We conclude that microbiome sequencing is useful to characterize complex food microbial communities, while additional work is required for predicting specific species viability from total RNA sequencing.

microbiology↗

bioRxiv2020

Genetic control of Campylobacter colonisation in broiler chickens: genomic and transcriptomic characterisation

Campylobacter is the leading cause of bacterial foodborne gastroenteritis in many countries. Source attribution studies unequivocally identify the handling or consumption of contaminated poultry meat as the primary risk factor. One potential strategy to control Campylobacter is to select poultry with increased resistance to colonisation. We conducted genomic and transcriptomic analyses of commercial pedigree broilers exposed to Campylobacter to examine persistent colonisation of the caecum as a quantitative trait. 3,000 broilers were genotyped using a 50K single nucleotide polymorphism (SNP) array and imputed to 600K SNPs. Genotypes were analysed for associations with the number of viable Campylobacter in the caeca. Heritability of the trait was modest but significantly greater than zero (h2=0.11 {+/-} 0.03). Genome-wide association analyses confirmed quantitative trait loci (QTL) on chromosomes 14 and 16 previously identified using the progeny of crosses of inbred lines differing in resistance, and detected two additional genome-wide significant QTLs on chromosomes 19 and 26. RNA-Seq analysis of the transcriptome of caecal tonsils from birds at the low and high extremes of C. jejuni colonisation phenotype identified differentially transcribed genes, mainly located within the QTL on chromosome 16 and proximal to the major histocompatibility complex (MHC) locus. We also identified strong cis-QTLs located within the MHC suggesting the presence of cis-acting variation in both MHC class I, class II and BG genes. Multiple other cis-acting variants were identified in association with key immune genes (COPS3, CCL4, CR1L, C4BP, PLGR) in the other QTLs. Pathway and network analysis implicated cooperative functional pathways and networks in colonisation, including those related to antigen presentation, innate and adaptive immune responses, calcium, and renin-angiotensin signalling. While co-selection for enhanced resistance and other breeding goal traits is feasible, the frequency of resistance-associated alleles was high in the population studied and non-genetic factors significantly influence Campylobacter colonisation in poultry. Author summaryCampylobacter infection is estimated to cause 95 million illnesses in people worldwide each year. Human infections mostly involve gastroenteritis, but can have severe complications. The handling or consumption of contaminated poultry meat is a key risk factor for human campylobacteriosis. The bacteria reach high numbers in the intestines of chickens reared for meat (broilers) and are frequently found on carcasses after slaughter. Effective vaccines against Campylobacter are not yet available, and treatments to reduce carcass contamination (e.g. chlorination) are not acceptable in some markets. One alternative is to breed for chickens with improved resistance to Campylobacter colonisation. To test the feasibility of this option in commercial birds, we analysed the genetic make-up of 3,000 pedigree broilers and determined the number of Campylobacter in their gut. There were associations between specific regions of the chicken genome and resistance to Campylobacter. Within some of these regions, expression of certain genes differed between birds at the low and high extremes of Campylobacter colonisation, providing a potential explanation for genetic variation in resistance. Selection of poultry with increased resistance to Campylobacter colonisation may be a complementary strategy to improved biosecurity, management, handling and processing procedures to reduce the burden of Campylobacter on human health.

genomics↗

bioRxiv2020

Integrative and Conjugative Elements (ICE) and Associated Cargo Genes within and across Hundreds of Bacterial Genera

Horizontal gene transfer mediated by integrative and conjugative elements (ICE) is considered an important evolutionary mechanism of bacteria. It allows organisms to quickly evolve new phenotypic properties including antimicrobial resistance (AMR) and virulence. The rate of ICE-mediated cargo gene exchange has not yet been comprehensively studied within and between bacterial taxa. In this paper we report a big data analysis of ICE and associated cargo genes across over 200,000 bacterial genomes representing 1,345 genera. Our results reveal that half of bacterial genomes contain one or more known ICE features ("ICE genomes"), and that the associated genetic cargo may play an important role in the spread of AMR genes within and between bacterial genera. We identify 43 AMR genes that appear only in ICE genomes and never in non-ICE genomes. A further set of 95 AMR genes are found >5x more often in ICE versus non-ICE genomes. In contrast, only 29 AMR genes are observed more frequently (at least 5:1) in non-ICE genomes compared to ICE genomes. Analysis of NCBI antibiotic susceptibility assay data reveals that ICE genomes are also over-represented amongst phenotypically resistant isolates, suggesting that ICE processes are critical for both genotypic and phenotypic AMR. These results, as well as the underlying big data resource, are important foundational tools for understanding bacterial evolution, particularly in relation to important bacterial phenotypes such as AMR.

microbiology↗

bioRxiv2020

Resources for genome editing in livestock: Cas9-expressing chickens and pigs

Genetically modified animals continue to provide important insights in biomedical sciences. Research has focused mostly on genetically modified mice so far, but other species like pigs resemble more closely the human physiology. In addition, cross-species comparisons with phylogenetically distant species such as chickens provide powerful insights into fundamental biological and biomedical processes. One of the most versatile genetic methods applicable across species is CRISPR/Cas9. Here, we report for the first time the generation of Cas9 transgenic chickens and pigs that allow in vivo genome editing in these two important agricultural species. We demonstrated that Cas9 is constitutively expressed in all organs of both species and that the animals are healthy and fertile. In addition, we confirmed the functionality of Cas9 for a number of different target genes and for a variety of cell types. Taken together, these transgenic animal species expressing Cas9 provide an unprecedented tool for agricultural and biomedical research, and will facilitate organ specific reverse genetics as well as cross-species comparisons. Significance statementGenome engineering of animals is crucial for translational medicine and the study of genetic traits. Here, we generated transgenic chickens and pigs that ubiquitously express the Cas9 endonuclease, providing the basis for in vivo genome editing. We demonstrated the functionality of this system by successful genome editing in chicken and porcine cells and tissues. These animals facilitate organ specific in vivo genome editing in both species without laborious germ line modifications, which will reduce the number of animals needed for genetic studies. They also provide a new tool for functional genomics, developmental biology and numerous other applications in biomedical and agricultural science.

molecular biology↗

bioRxiv2019

Survey of Public Assay Data: Opportunities and Challenges to Understanding Antimicrobial Resistance

ABSTRACTComputational learning methods allow researchers to make predictions, draw inferences, and automate generation of mathematical models. These models are crucial to solving real world problems, such as antimicrobial resistance, pathogen detection, and protein evolution. Machine learning methods depend upon ground truth data to achieve specificity and sensitivity. Since the data is limited in this case, as we will show during the course of this paper, and as the size of available data increases super-linearly, it is of paramount importance to understand the distribution of ground truth data and the analyses it is suited and where it may have limitations that bias downstream learning methods. In this paper, we focus on training data required to model antimicrobial resistance (AR). We report an analysis of bacterial biochemical assay data associated with whole genome sequencing (WGS) from the National Center for Biotechnology Information (NCBI), and discuss important implications when making use of assay data, utilizing genetic features as training data for machine learning models. Complete discussion of machine learning model implementation is outside the scope of this paper and the subject to a later publication. The antimicrobial assay data was obtained from NCBI BioSample, which contains descriptive information about the physical biological specimen from which experimental data is obtained and the results of those experiments themselves.[1] Assay data includes minimum inhibitory concentrations (MIC) of antibiotics, links to associated microbial WGS data, and treatment of a particular microorganism with antibiotics. We observe that there is minimal microbial data available for many antibiotics and for targeted taxonomic groups. The antibiotics with the highest number of assays have less than 1500 measurements each. Corresponding bias in available assays makes machine learning problematic for some important microbes and for building more advanced models that can work across microbial genera. In this study we focus, therefore, on the antibiotic with most assay data (tetracycline) and the corresponding genus with the most available sequence (Acinetobacter with 14000 measurements across 49 antibiotic compounds). Using this data for training and testing, we observed contradictions in the distribution of assay outcomes and report methods to identify and resolve such conflicts. Per antibiotic, we find that there can be up to 30% of (resolvable) conflicting measurements. As more data becomes available, automated training data curation will be an important part of creating useful machine learning models to predict antibiotic resistance. CCS CONCEPTS* Applied computing [->] Computational biology; Computational genomics; Bioinformatics;

bioinformatics↗

bioRxiv2018

Supporting decision making: Modeling and forecasting measles in a London borough

To investigate the feasibility of using freeware to model and forecast disease on a local scale, we report the results of modeling measles using a spatial patch model centered around 73 clinics in the North West London Borough of Ealing. MMR1 and MM2 immunization data was extracted for three cohorts, age 1-3, 4-6 and 7-19 and patient population was estimated using general practice profile records.\n\nWe designed the measles model using the open source Spatiotemporal Epidemiological Modeler (STEM), extending a compartmental disease model to include both maternal immunity and delays in antibody response after immunization. Individuals above age 19 are not included in the modeling.\n\nNext, we generate an approximate 20-year model of vaccination coverage for Ealing. In England, children are immunized between age 1 and 2, then again at around age 5; hence immunization events are modeled for the age 1-3 and age 4-6 cohorts. Parameter values were based on measles research literature; transmission coefficients were estimated using the Polymod contact data and also fitted to 2011-2012 case reporting data for Ealing.\n\nTo examine possible effects of policy change, we create two scenarios A and B. In A, we increase vaccination coverage by 10% for all clinics; in B, we focus only on the bottom 10% of the poorest performing clinics (8 clinics total) and equivalently improve their coverage. Scenario A reduces measles from an initial level of 60 cases per year (2011) to 26 cases per year in 2017 (a 58% reduction), compared to the status quo which declines to 45 cases per year. Scenario B reduces measles by 44%, or to 34 cases per year in 2017.\n\nWe conclude that local scale modeling is possible, and that the transparency of analysis provided by an open source application lends credence to the output of the models.

epidemiology↗