Dataset Information

A generalizable 29-mRNA neural-network classifier for acute bacterial and viral infections.

ABSTRACT: Improved identification of bacterial and viral infections would reduce morbidity from sepsis, reduce antibiotic overuse, and lower healthcare costs. Here, we develop a generalizable host-gene-expression-based classifier for acute bacterial and viral infections. We use training data (N?=?1069) from 18 retrospective transcriptomic studies. Using only 29 preselected host mRNAs, we train a neural-network classifier with a bacterial-vs-other area under the receiver-operating characteristic curve (AUROC) 0.92 (95% CI 0.90-0.93) and a viral-vs-other AUROC 0.92 (95% CI 0.90-0.93). We then apply this classifier, inflammatix-bacterial-viral-noninfected-version 1 (IMX-BVN-1), without retraining, to an independent cohort (N?=?163). In this cohort, IMX-BVN-1 AUROCs are: bacterial-vs.-other 0.86 (95% CI 0.77-0.93), and viral-vs.-other 0.85 (95% CI 0.76-0.93). In patients enrolled within 36?h of hospital admission (N?=?70), IMX-BVN-1 AUROCs are: bacterial-vs.-other 0.92 (95% CI 0.83-0.99), and viral-vs.-other 0.91 (95% CI 0.82-0.98). With further study, IMX-BVN-1 could provide a tool for assessing patients with suspected infection and sepsis at hospital admission.

SUBMITTER: Mayhew MB

PROVIDER: S-EPMC7055276 | biostudies-literature | 2020 Mar

REPOSITORIES: biostudies-literature

ACCESS DATA

Publications

A generalizable 29-mRNA neural-network classifier for acute bacterial and viral infections.

Mayhew Michael B MB Buturovic Ljubomir L Luethy Roland R Midic Uros U Moore Andrew R AR Roque Jonasel A JA Shaller Brian D BD Asuni Tola T Rawling David D Remmel Melissa M Choi Kirindi K Wacker James J Khatri Purvesh P Rogers Angela J AJ Sweeney Timothy E TE

Nature communications 20200304 1

Improved identification of bacterial and viral infections would reduce morbidity from sepsis, reduce antibiotic overuse, and lower healthcare costs. Here, we develop a generalizable host-gene-expression-based classifier for acute bacterial and viral infections. We use training data (N = 1069) from 18 retrospective transcriptomic studies. Using only 29 preselected host mRNAs, we train a neural-network classifier with a bacterial-vs-other area under the receiver-operating characteristic curve (AUR ...[more]

PMID: 32132525

Similar Datasets

Project description:BackgroundWhether or not to administer antibiotics is a common and challenging clinical decision in patients with suspected infections presenting to the emergency department (ED). We prospectively validate InSep, a 29-mRNA blood-based host response test for the prediction of bacterial and viral infections.MethodsThe PROMPT trial is a prospective, non-interventional, multi-center clinical study that enrolled 397 adult patients presenting to the ED with signs of acute infection and at least one vital sign change. The infection status was adjudicated using chart review (including a syndromic molecular respiratory panel, procalcitonin and C-reactive protein) by three infectious disease physicians blinded to InSep results. InSep (version BVN-2) was performed using PAXgene Blood RNA processed and quantified on NanoString nCounter SPRINT. InSep results (likelihood of bacterial and viral infection) were compared to the adjudicated infection status.ResultsSubject mean age was 64 years, comorbidities were significant for diabetes (17.1%), chronic obstructive pulmonary disease (13.6%), and severe neurological disease (6.8%); 16.9% of subjects were immunocompromised. Infections were adjudicated as bacterial (14.1%), viral (11.3%) and noninfected (0.25%): 74.1% of subjects were adjudicated as indeterminate. InSep distinguished bacterial vs. viral/noninfected patients and viral vs. bacterial/noninfected patients using consensus adjudication with AUROCs of 0.94 (95% CI 0.90-0.99) and 0.90 (95% CI 0.83-0.96), respectively. AUROCs for bacterial vs. viral/noninfected patients were 0.88 (95% CI 0.79-0.96) for PCT, 0.80 (95% CI 0.72-89) for CRP and 0.78 (95% CI 0.69-0.87) for white blood cell counts (of note, the latter biomarkers were provided as part of clinical adjudication). To enable clinical actionability, InSep incorporates score cutoffs to allocate patients into interpretation bands. The Very Likely (rule in) InSep bacterial band showed a specificity of 98% compared to 94% for the corresponding PCT band (> 0.5 µg/L); the Very Unlikely (rule-out) band showed a sensitivity of 95% for InSep compared to 86% for PCT. For the detection of viral infections, InSep demonstrated a specificity of 93% for the Very Likely band (rule in) and a sensitivity of 96% for the Very Unlikely band (rule out).ConclusionsInSep demonstrated high accuracy for predicting the presence of both bacterial and viral infections in ED patients with suspected acute infections or suspected sepsis. When translated into a rapid, point-of-care test, InSep will provide ED physicians with actionable results supporting early informed treatment decisions to improve patient outcomes while upholding antimicrobial stewardship. Registration number at Clinicaltrials.gov NCT03295825.

Project description:Concussion is a global health concern. Despite its high prevalence, a sound understanding of the mechanisms underlying this type of diffuse brain injury remains elusive. It is, however, well established that concussions cause significant functional deficits; that children and youths are disproportionately affected and have longer recovery time than adults; and that individuals suffering from a concussion are more prone to experience additional concussions, with each successive injury increasing the risk of long term neurological and mental health complications. Currently, the most significant challenge in concussion management is the lack of objective, clinically- accepted, brain-based approaches for determining whether an athlete has suffered a concussion. Here, we report on our efforts to address this challenge. Specifically, we introduce a deep learning long short-term memory (LSTM)-based recurrent neural network that is able to distinguish between non-concussed and acute post-concussed adolescent athletes using only short (i.e. 90 s long) samples of resting state EEG data as input. The athletes were neither required to perform a specific task nor expected to respond to a stimulus during data collection. The acquired EEG data were neither filtered, cleaned of artefacts, nor subjected to explicit feature extraction. The LSTM network was trained and validated using data from 27 male, adolescent athletes with sports related concussion, benchmarked against 35 non-concussed adolescent athletes. During rigorous testing, the classifier consistently identified concussions with an accuracy of > 90% and achieved an ensemble median Area Under the Receiver Operating Characteristic Curve (ROC/AUC) equal to 0.971. This is the first instance of a high-performing classifier that relies only on easy-to-acquire resting state, raw EEG data. Our concussion classifier represents a promising first step towards the development of an easy-to-use, objective, brain-based, automatic classification of concussion at an individual level.

Project description:Bacterial and viral infections are often clinically indistinguishable, leading to inappropriate patient management and antibiotic misuse. Bacterial-induced host proteins such as procalcitonin, C-reactive protein (CRP), and Interleukin-6, are routinely used to support diagnosis of infection. However, their performance is negatively affected by inter-patient variability, including time from symptom onset, clinical syndrome, and pathogens. Our aim was to identify novel viral-induced host proteins that can complement bacterial-induced proteins to increase diagnostic accuracy. Initially, we conducted a bioinformatic screen to identify putative circulating host immune response proteins. The resulting 600 candidates were then quantitatively screened for diagnostic potential using blood samples from 1002 prospectively recruited patients with suspected acute infectious disease and controls with no apparent infection. For each patient, three independent physicians assigned a diagnosis based on comprehensive clinical and laboratory investigation including PCR for 21 pathogens yielding 319 bacterial, 334 viral, 112 control and 98 indeterminate diagnoses; 139 patients were excluded based on predetermined criteria. The best performing host-protein was TNF-related apoptosis-inducing ligand (TRAIL) (area under the curve [AUC] of 0.89; 95% confidence interval [CI], 0.86 to 0.91), which was consistently up-regulated in viral infected patients. We further developed a multi-protein signature using logistic-regression on half of the patients and validated it on the remaining half. The signature with the highest precision included both viral- and bacterial-induced proteins: TRAIL, Interferon gamma-induced protein-10, and CRP (AUC of 0.94; 95% CI, 0.92 to 0.96). The signature was superior to any of the individual proteins (P<0.001), as well as routinely used clinical parameters and their combinations (P<0.001). It remained robust across different physiological systems, times from symptom onset, and pathogens (AUCs 0.87-1.0). The accurate differential diagnosis provided by this novel combination of viral- and bacterial-induced proteins has the potential to improve management of patients with acute infections and reduce antibiotic misuse.

Project description:BackgroundDealing with the high dimension of both neuroimaging data and genetic data is a difficult problem in the association of genetic data to neuroimaging. In this article, we tackle the latter problem with an eye toward developing solutions that are relevant for disease prediction. Supported by a vast literature on the predictive power of neural networks, our proposed solution uses neural networks to extract from neuroimaging data features that are relevant for predicting Alzheimer's Disease (AD) for subsequent relation to genetics. The neuroimaging-genetic pipeline we propose is comprised of image processing, neuroimaging feature extraction and genetic association steps. We present a neural network classifier for extracting neuroimaging features that are related with the disease. The proposed method is data-driven and requires no expert advice or a priori selection of regions of interest. We further propose a multivariate regression with priors specified in the Bayesian framework that allows for group sparsity at multiple levels including SNPs and genes.ResultsWe find the features extracted with our proposed method are better predictors of AD than features used previously in the literature suggesting that single nucleotide polymorphisms (SNPs) related to the features extracted by our proposed method are also more relevant for AD. Our neuroimaging-genetic pipeline lead to the identification of some overlapping and more importantly some different SNPs when compared to those identified with previously used features.ConclusionsThe pipeline we propose combines machine learning and statistical methods to benefit from the strong predictive performance of blackbox models to extract relevant features while preserving the interpretation provided by Bayesian models for genetic association. Finally, we argue in favour of using automatic feature extraction, such as the method we propose, in addition to ROI or voxelwise analysis to find potentially novel disease-relevant SNPs that may not be detected when using ROIs or voxels alone.

Dataset Information

A generalizable 29-mRNA neural-network classifier for acute bacterial and viral infections.

Publications

A generalizable 29-mRNA neural-network classifier for acute bacterial and viral infections.

Similar Datasets

OmicsDI is part of the ELIXIR infrastructure

Tweets