RESUMO
BACKGROUND: Intravenous (IV) fluid contamination is a common cause of preanalytical error that can delay or misguide treatment decisions, leading to patient harm. Current approaches for detecting contamination rely on delta checks, which require a prior result, or manual technologist intervention, which is inefficient and vulnerable to human error. Supervised machine learning may provide a means to detect contamination, but its implementation is hindered by its reliance on expert-labeled training data. An automated approach that is accurate, reproducible, and practical is needed. METHODS: A total of 25 747 291 basic metabolic panel (BMP) results from 312 721 patients were obtained from the laboratory information system (LIS). A Uniform Manifold Approximation and Projection (UMAP) model was trained and tested using a combination of real patient data and simulated IV fluid contamination. To provide an objective metric for classification, an "enrichment score" was derived and its performance assessed. Our current workflow was compared to UMAP predictions using expert chart review. RESULTS: UMAP embeddings from real patient results demonstrated outliers suspicious for IV fluid contamination when compared with the simulated contamination's embeddings. At a flag rate of 3 per 1000 results, the positive predictive value (PPV) was adjudicated to be 0.78 from 100 consecutive positive predictions. Of these, 58 were previously undetected by our current clinical workflows, with 49 BMPs displaying a total of 56 critical results. CONCLUSIONS: Accurate and automatable detection of IV fluid contamination in BMP results is achievable without curating expertly labeled training data.
Assuntos
Aprendizado de Máquina não Supervisionado , Humanos , Valor Preditivo dos Testes , Fluxo de TrabalhoRESUMO
BACKGROUND: Diagnosing acute kidney injury (AKI) and chronic kidney disease (CKD) relies on creatinine, which lacks optimal diagnostic sensitivity. The kidney-specific proximal tubular enzyme myo-inositol oxygenase (MIOX) catalyzes the conversion of myo-inositol (MI) to D-glucuronic acid. We hypothesized that proximal tubular damage, which occurs in AKI and CKD, will decrease MIOX activity, causing MI accumulation. To explore this, we developed an LC-MS/MS assay to quantify plasma MI and assessed its potential in identifying AKI and CKD patients. METHODS: MI was quantified in plasma from 3 patient cohorts [normal kidney function (n = 105), CKD (n = 94), and AKI (n = 54)]. The correlations between MI and creatinine were determined using Deming regression and Pearson correlation and the impact of age, sex, and ethnicity on MI concentrations was assessed. Receiver operating characteristic curve analysis was employed to evaluate MI diagnostic performance. RESULTS: In volunteers with normal kidney function, the central 95th percentile range of plasma MI concentrations was 16.6 to 44.2 µM. Age, ethnicity, and sex showed minimal influence on MI. Patients with AKI and CKD exhibited higher median MI concentrations [71.1 (25th percentile: 38.2, 75th percentile: 115.4) and 102.4 (77, 139.5) µM], respectively. MI exhibited excellent sensitivity (98.9%) and specificity (100%) for diagnosing CKD. In patients with AKI, MI increased 32.9 (SD 16.8) h before creatinine. CONCLUSIONS: This study unveils MI as a potential renal biomarker, notably elevated in plasma during AKI and CKD. Plasma MI rises 33 h prior to serum creatinine, enabling early AKI detection. Further validation and exploration of MI quantitation in kidney disease diagnosis is warranted.
RESUMO
BACKGROUND: Measuring parathyroid hormone-related peptide (PTHrP) helps diagnose the humoral hypercalcemia of malignancy, but is often ordered for patients with low pretest probability, resulting in poor test utilization. Manual review of results to identify inappropriate PTHrP orders is a cumbersome process. METHODS: Using a dataset of 1330 patients from a single institute, we developed a machine learning (ML) model to predict abnormal PTHrP results. We then evaluated the performance of the model on two external datasets. Different strategies (model transporting, retraining, rebuilding, and fine-tuning) were investigated to improve model generalizability. Maximum mean discrepancy (MMD) was adopted to quantify the shift of data distributions across different datasets. RESULTS: The model achieved an area under the receiver operating characteristic curve (AUROC) of 0.936, and a specificity of 0.842 at 0.900 sensitivity in the development cohort. Directly transporting this model to two external datasets resulted in a deterioration of AUROC to 0.838 and 0.737, with the latter having a larger MMD corresponding to a greater data shift compared to the original dataset. Model rebuilding using site-specific data improved AUROC to 0.891 and 0.837 on the two sites, respectively. When external data is insufficient for retraining, a fine-tuning strategy also improved model utility. CONCLUSIONS: ML offers promise to improve PTHrP test utilization while relieving the burden of manual review. Transporting a ready-made model to external datasets may lead to performance deterioration due to data distribution shift. Model retraining or rebuilding could improve generalizability when there are enough data, and model fine-tuning may be favorable when site-specific data is limited.
Assuntos
Hipercalcemia , Neoplasias , Humanos , Proteína Relacionada ao Hormônio Paratireóideo , Curva ROC , Aprendizado de MáquinaRESUMO
BACKGROUND: Serum free light chain (sFLC) assays are interpreted using a sFLC-ratio-based reference interval (manufacturer's interval) that was defined using a cohort of healthy patients. However, renal impairment elevates the sFLC-ratio, leading to a high false positive rate when using the manufacturer's interval. Prior studies have developed renal-specific reference intervals; however, this approach has not been widely adopted due to practical limitations. Thus, there remains a critical need for a renally robust sFLC interpretation method. METHODS: Retrospective data mining was used to define patient cohorts that reflect the spectrum of renal function seen in clinical practice. Two new reference intervals, one based on the sFLC-ratio and one based on a novel principal component analysis (PCA)-based metric, were developed for the FREELITE assay (Binding Site) on the Roche Cobas c501 instrument (Roche). RESULTS: Compared to the manufacturer's reference interval, both new methods exhibited significantly lower false positive rates and greater robustness to renal function while maintaining equivalent sensitivity for monoclonal gammopathy (MG) diagnosis. While not significantly different, the point estimate for sensitivity was highest for the PCA-based approach. CONCLUSION: Renally robust sFLC interpretation using a single reference interval is possible given a reference cohort that reflects the variation in renal function observed in practice. Further studies are needed to achieve sufficient power and determine if the novel PCA-based metric offers superior sensitivity for MG diagnosis. These new methods offer the practical advantages of not requiring an estimated glomerular filtration rate result or multiple reference intervals, thereby lowering practical barriers to implementation.
RESUMO
The Pseudomonas syringae DC3000 type III effector HopAM1 suppresses plant immunity and contains a Toll/interleukin-1 receptor (TIR) domain homologous to immunity-related TIR domains of plant nucleotide-binding leucine-rich repeat receptors that hydrolyze nicotinamide adenine dinucleotide (NAD+ ) and activate immunity. In vitro and in vivo assays were conducted to determine if HopAM1 hydrolyzes NAD+ and if the activity is essential for HopAM1's suppression of plant immunity and contribution to virulence. HPLC and LC-MS were utilized to analyze metabolites produced from NAD+ by HopAM1 in vitro and in both yeast and plants. Agrobacterium-mediated transient expression and in planta inoculation assays were performed to determine HopAM1's intrinsic enzymatic activity and virulence contribution. HopAM1 is catalytically active and hydrolyzes NAD+ to produce nicotinamide and a novel cADPR variant (v2-cADPR). Expression of HopAM1 triggers cell death in yeast and plants dependent on the putative catalytic residue glutamic acid 191 (E191) within the TIR domain. Furthermore, HopAM1's E191 residue is required to suppress both pattern-triggered immunity and effector-triggered immunity and promote P. syringae virulence. HopAM1 manipulates endogenous NAD+ to produce v2-cADPR and promote pathogenesis. This work suggests that HopAM1's TIR domain possesses different catalytic specificity than other TIR domain-containing NAD+ hydrolases and that pathogens exploit this activity to sabotage NAD+ metabolism for immune suppression and virulence.
Assuntos
Proteínas de Arabidopsis , Arabidopsis , Arabidopsis/metabolismo , Proteínas de Arabidopsis/metabolismo , Proteínas de Bactérias/metabolismo , NAD/metabolismo , Doenças das Plantas/microbiologia , Pseudomonas syringae/fisiologia , Receptores de Interleucina-1/metabolismo , VirulênciaRESUMO
OBJECTIVE: Clinicians order laboratory tests in an effort to reduce diagnostic or therapeutic uncertainty. Information theory provides the opportunity to quantify the degree to which a test result is expected to reduce diagnostic uncertainty. We sought to apply information theory toward the evaluation and optimization of a diagnostic test threshold and to determine if the results would differ from those of conventional methodologies. We used a heparin/PF4 immunoassay (PF4 ELISA) as a case study. MATERIALS AND METHODS: The laboratory database was queried for PF4 ELISA and serotonin release assay (SRA) results during the study period, with the latter serving as the gold standard for the disease heparin-induced thrombocytopenia (HIT). The optimized diagnostic threshold of the PF4 ELISA test was compared using conventional versus information theoretic approaches under idealized (pretest probability = 50%) and realistic (pretest probability = 2.4%) testing conditions. RESULTS: Under ideal testing conditions, both analyses yielded a similar optimized optical density (OD) threshold of OD > 0.79. Under realistic testing conditions, information theory suggested a higher threshold, OD > 1.5 versus OD > 0.6. Increasing the diagnostic threshold improved the global information value, the value of a positive test and the noise content with only a minute change in the negative test value. DISCUSSION: Our information theoretic approach suggested that the current FDA approved cutoff (OD > 0.4) is overly permissive leading to loss of test value and injection of noise into an already complex diagnostic dilemma. Because our approach is purely statistical and takes as input data that are readily accessible in the clinical laboratory it offers a scalable and data-driven strategy for optimizing test value that may be widely applicable in the domain of laboratory medicine. CONCLUSION: Information theory provides more meaningful measures of test value than the widely used accuracy-based metrics.
Assuntos
Médicos , Trombocitopenia , Heparina/efeitos adversos , Humanos , Teoria da Informação , Fator Plaquetário 4RESUMO
Voltage-gated ion channels generate dynamic ionic currents that are vital to the physiological functions of many tissues. These proteins contain separate voltage-sensing domains, which detect changes in transmembrane voltage, and pore domains, which conduct ions. Coupling of voltage sensing and pore opening is critical to the channel function and has been modeled as a protein-protein interaction between the two domains. Here, we show that coupling in Kv7.1 channels requires the lipid phosphatidylinositol 4,5-bisphosphate (PIP2). We found that voltage-sensing domain activation failed to open the pore in the absence of PIP2. This result is due to loss of coupling because PIP2 was also required for pore opening to affect voltage-sensing domain activation. We identified a critical site for PIP2-dependent coupling at the interface between the voltage-sensing domain and the pore domain. This site is actually a conserved lipid-binding site among different K(+) channels, suggesting that lipids play an important role in coupling in many ion channels.
Assuntos
Ativação do Canal Iônico/fisiologia , Canal de Potássio KCNQ1/metabolismo , Modelos Biológicos , Fosfatidilinositol 4,5-Difosfato/metabolismo , Algoritmos , Sequência de Aminoácidos , Animais , Sítios de Ligação/genética , Western Blotting , Feminino , Humanos , Ativação do Canal Iônico/genética , Canal de Potássio KCNQ1/química , Canal de Potássio KCNQ1/genética , Potenciais da Membrana/genética , Potenciais da Membrana/fisiologia , Modelos Moleculares , Dados de Sequência Molecular , Mutação , Oócitos/metabolismo , Oócitos/fisiologia , Técnicas de Patch-Clamp , Fosfatidilinositol 4,5-Difosfato/química , Ligação Proteica , Estrutura Terciária de Proteína , Homologia de Sequência de Aminoácidos , Xenopus laevisRESUMO
Phosphatidylinositol 4,5-bisphosphate (PIP(2)) is necessary for the function of various ion channels. The potassium channel, I(Ks), is important for cardiac repolarization and requires PIP(2) to activate. Here we show that the auxiliary subunit of I(Ks), KCNE1, increases PIP(2) sensitivity 100-fold over channels formed by the pore-forming KCNQ1 subunits alone, which effectively amplifies current because native PIP(2) levels in the membrane are insufficient to activate all KCNQ1 channels. A juxtamembranous site in the KCNE1 C terminus is a key structural determinant of PIP(2) sensitivity. Long QT syndrome associated mutations of this site lower PIP(2) affinity, resulting in reduced current. Application of exogenous PIP(2) to these mutants restores wild-type channel activity. These results reveal a vital role of PIP(2) for KCNE1 modulation of I(Ks) channels that may represent a common mechanism of auxiliary subunit modulation of many ion channels.
Assuntos
Canal de Potássio KCNQ1/metabolismo , Mutação , Fosfatidilinositol 4,5-Difosfato/metabolismo , Canais de Potássio de Abertura Dependente da Tensão da Membrana/metabolismo , Proteínas de Xenopus/metabolismo , Sequência de Aminoácidos , Animais , Humanos , Íons , Lipídeos/química , Síndrome do QT Longo/genética , Dados de Sequência Molecular , Técnicas de Patch-Clamp , Canais de Potássio/química , Estrutura Terciária de Proteína , Homologia de Sequência de Aminoácidos , Xenopus laevisRESUMO
Molecular microbiology assays have a higher cost of testing compared to traditional methods and need to be utilized appropriately. Results from these assays may also require interpretation and appropriate follow-up. Electronic tools available in the electronic health record and laboratory information system can be deployed both preanalytically and postanalytically to influence ordering behaviors and positively impact diagnostic stewardship. Next generation technologies, such as machine learning and artificial intelligence, have the potential to expand upon the capabilities currently available and warrant additional study and development but also require regulation around their use in health care.
Assuntos
Sistemas de Informação em Laboratório Clínico , Registros Eletrônicos de Saúde , Inteligência ArtificialRESUMO
BACKGROUND: Analysis of whole blood specimens is rapid and saves blood, but hemolysis may go undetected and compromise the accuracy of potassium measurement. We aimed to define the frequency and magnitude of error in whole blood potassium measurement. METHODS: 34 months of whole blood and plasma potassium data were extracted from patients aged less than 2 years at the time of sample acquisition. Hemolysis was detected using the plasma "H index." The magnitude of potassium bias was estimated from the difference between paired whole blood and plasma measurement separated by less than 2 h. RESULTS: 56,000 of the 105,000 data points were from plasma and 20 % of these had significant hemolysis. Rates of hemolysis (nearing 50 %) were greatest in the neonatal nursery. Of 662 proximal whole blood and plasma paired results, 8 % had elevated whole blood potassium with a normal plasma value and 4 % had a normal whole blood potassium with reduced plasma potassium. The bias between whole blood and plasma potassium ranged from -1.0 to 4.0 mmol/L. CONCLUSIONS: The use of whole blood analysis brings with it significant risk for error in potassium measurement. Better tools to detect hemolysis in these types of specimens are indicated.
Assuntos
Hemólise , Potássio , Recém-Nascido , Humanos , Criança , Testes Hematológicos , Valores de ReferênciaAssuntos
Alérgenos/imunologia , Hipersensibilidade Alimentar/diagnóstico , Hipersensibilidade Alimentar/imunologia , Imunoglobulina E/imunologia , Carne Vermelha/efeitos adversos , Especificidade de Anticorpos/imunologia , Ensaio de Imunoadsorção Enzimática , Hipersensibilidade Alimentar/sangue , Humanos , Imunoglobulina E/sangue , Sensibilidade e Especificidade , Testes CutâneosRESUMO
BACKGROUND: Methods of machine learning provide opportunities to use real-world data to solve complex problems. Applications of these methods in laboratory medicine promise to increase diagnostic accuracy and streamline laboratory operations leading to improvement in the quality and efficiency of healthcare delivery. However, machine learning models are vulnerable to learning from undesirable patterns in the data that reflect societal biases. As a result, irresponsible application of machine learning may lead to the perpetuation, or even amplification, of existing disparities in healthcare outcomes. CONTENT: In this work, we review what it means for a model to be unfair, discuss the various ways that machine learning models become unfair, and present engineering principles emerging from the field of algorithmic fairness. These materials are presented with a focus on the development of machine learning models in laboratory medicine. SUMMARY: We hope that this work will serve to increase awareness, and stimulate further discussion, of this important issue among laboratorians as the field moves forward with the incorporation of machine learning models into laboratory practice.
Assuntos
Atenção à Saúde , Aprendizado de Máquina , Humanos , Algoritmos , Laboratórios , ViésRESUMO
Cellular behaviors emerge from layers of molecular interactions: proteins interact to form complexes, pathways, and phenotypes. We show that hierarchical networks of protein interactions can be defined from the statistical pattern of proteome variation measured across thousands of diverse bacteria and that these networks reflect the emergence of complex bacterial phenotypes. Our results are validated through gene-set enrichment analysis and comparison to existing experimentally derived databases. We demonstrate the biological utility of our approach by creating a model of motility in Pseudomonas aeruginosa and using it to identify a protein that affects pilus-mediated motility. Our method, SCALES (Spectral Correlation Analysis of Layered Evolutionary Signals), may be useful for interrogating genotype-phenotype relationships in bacteria.
Assuntos
Mapas de Interação de Proteínas , Proteoma , Bactérias/genética , Fímbrias Bacterianas , FenótipoRESUMO
Cyclic adenosine diphosphate (ADP)-ribose (cADPR) isomers are signaling molecules produced by bacterial and plant Toll/interleukin-1 receptor (TIR) domains via nicotinamide adenine dinucleotide (oxidized form) (NAD+) hydrolysis. We show that v-cADPR (2'cADPR) and v2-cADPR (3'cADPR) isomers are cyclized by O-glycosidic bond formation between the ribose moieties in ADPR. Structures of 2'cADPR-producing TIR domains reveal conformational changes that lead to an active assembly that resembles those of Toll-like receptor adaptor TIR domains. Mutagenesis reveals a conserved tryptophan that is essential for cyclization. We show that 3'cADPR is an activator of ThsA effector proteins from the bacterial antiphage defense system termed Thoeris and a suppressor of plant immunity when produced by the effector HopAM1. Collectively, our results reveal the molecular basis of cADPR isomer production and establish 3'cADPR in bacteria as an antiviral and plant immunity-suppressing signaling molecule.
Assuntos
ADP-Ribosil Ciclase , Proteínas Adaptadoras de Transporte Vesicular , Bactérias , Proteínas de Bactérias , ADP-Ribose Cíclica , Imunidade Vegetal , Receptores Toll-Like , ADP-Ribosil Ciclase/química , ADP-Ribosil Ciclase/genética , ADP-Ribosil Ciclase/metabolismo , Proteínas Adaptadoras de Transporte Vesicular/química , Proteínas Adaptadoras de Transporte Vesicular/genética , Proteínas Adaptadoras de Transporte Vesicular/metabolismo , Bactérias/imunologia , Bactérias/virologia , Proteínas de Bactérias/química , Proteínas de Bactérias/genética , Proteínas de Bactérias/metabolismo , ADP-Ribose Cíclica/biossíntese , ADP-Ribose Cíclica/química , Isomerismo , NAD/metabolismo , Domínios Proteicos , Receptores de Interleucina-1/química , Transdução de Sinais , Receptores Toll-Like/química , Receptores Toll-Like/genética , Receptores Toll-Like/metabolismo , Triptofano/química , Triptofano/genéticaRESUMO
Protein domain-based approaches to analyzing sequence data are valuable tools for examining and exploring genomic architecture across genomes of different organisms. Here, we present a complete dataset of domains from the publicly available sequence data of 9,051 reference viral genomes. The data provided contain information such as sequence position and neighboring domains from 30,947 pHMM-identified domains from each reference viral genome. Domains were identified from viral whole-genome sequence using automated profile Hidden Markov Models (pHMM). This study also describes the framework for constructing "domain neighborhoods", as well as the dataset representing it. These data can be used to examine shared and differing domain architectures across viral genomes, to elucidate potential functional properties of genes, and potentially to classify viruses.
Assuntos
Bases de Dados de Proteínas , Genoma Viral , Domínios Proteicos , Cadeias de MarkovRESUMO
In voltage-gated potassium (KV) channels, the voltage-sensing domain (VSD) undergoes sequential activation from the resting state to the intermediate state and activated state to trigger pore opening via electro-mechanical (E-M) coupling. However, the spatial and temporal details underlying E-M coupling remain elusive. Here, utilizing KV7.1's unique two open states, we report a two-stage E-M coupling mechanism in voltage-dependent gating of KV7.1 as triggered by VSD activations to the intermediate and then activated state. When the S4 segment transitions to the intermediate state, the hand-like C-terminus of the VSD-pore linker (S4-S5L) interacts with the pore in the same subunit. When S4 then proceeds to the fully-activated state, the elbow-like hinge between S4 and S4-S5L engages with the pore of the neighboring subunit to activate conductance. This two-stage hand-and-elbow gating mechanism elucidates distinct tissue-specific modulations, pharmacology, and disease pathogenesis of KV7.1, and likely applies to numerous domain-swapped KV channels.