Acute febrile illness among outpatients seeking health care in Bangladeshi hospitals prior to the COVID-19 pandemic.

Understanding the distribution of pathogens causing acute febrile illness (AFI) is important for clinical management of patients in resource-poor settings. We evaluated the proportion of AFI caused by specific pathogens among outpatients in Bangladesh. During May 2019-March 2020, physicians screened patients aged ≥2 years in outpatient departments of four tertiary level public hospitals. We randomly enrolled patients having measured fever (≥100.4°F) during assessment with onset within the past 14 days. Blood and urine samples were tested at icddr,b through rapid diagnostic tests, bacterial culture, and polymerase chain reaction (PCR). Acute and convalescent samples were sent to the Centers for Disease Control and Prevention (USA) for Rickettsia and Orientia (R/O) and Leptospira tests. Among 690 patients, 69 (10%) had enteric fever (Salmonella enterica serotype Typhi orSalmonella enterica serotype Paratyphi), 51 (7.4%) Escherichia coli, and 28 (4.1%) dengue detected. Of the 441 patients tested for R/O, 39 (8.8%) had rickettsioses. We found 7 (2%) Leptospira cases among the 403 AFI patients tested. Nine patients (1%) were hospitalized, and none died. The highest proportion of enteric fever (15%, 36/231) and rickettsioses (14%, 25/182) was in Rajshahi. Dhaka had the most dengue cases (68%, 19/28). R/O affected older children and young adults (IQR 8-23 years) and was detected more frequently in the 21-25 years age-group (17%, 12/70). R/O was more likely to be found in patients in Rajshahi region than in Sylhet (aOR 2.49, 95% CI 0.85-7.32) between July and December (aOR 2.01, 1.01-5.23), and who had a history of recent animal entry inside their house than not (aOR 2.0, 0.93-4.3). Gram-negative Enterobacteriaceae were the most common bacterial infections, and dengue was the most common viral infection among AFI patients in Bangladeshi hospitals, though there was geographic variability. These results can help guide empiric outpatient AFI management.

Deep neural network for detecting arbitrary precision peptide features through attention based segmentation.

A promising technique of discovering disease biomarkers is to measure the relative protein abundance in multiple biofluid samples through liquid chromatography with tandem mass spectrometry (LC-MS/MS) based quantitative proteomics. The key step involves peptide feature detection in the LC-MS map, along with its charge and intensity. Existing heuristic algorithms suffer from inaccurate parameters and human errors. As a solution, we propose PointIso, the first point cloud based arbitrary-precision deep learning network to address this problem. It consists of attention based scanning step for segmenting the multi-isotopic pattern of 3D peptide features along with the charge, and a sequence classification step for grouping those isotopes into potential peptide features. PointIso achieves 98% detection of high-quality MS/MS identified peptide features in a benchmark dataset. Next, the model is adapted for handling the additional 'ion mobility' dimension and achieves 4% higher detection than existing algorithms on the human proteome dataset. Besides contributing to the proteomics study, our novel segmentation technique should serve the general object detection domain as well.

A draft genome sequence of the elusive giant squid, Architeuthis dux.

BACKGROUND: The giant squid (Architeuthis dux; Steenstrup, 1857) is an enigmatic giant mollusc with a circumglobal distribution in the deep ocean, except in the high Arctic and Antarctic waters. The elusiveness of the species makes it difficult to study. Thus, having a genome assembled for this deep-sea-dwelling species will allow several pending evolutionary questions to be unlocked. FINDINGS: We present a draft genome assembly that includes 200 Gb of Illumina reads, 4 Gb of Moleculo synthetic long reads, and 108 Gb of Chicago libraries, with a final size matching the estimated genome size of 2.7 Gb, and a scaffold N50 of 4.8 Mb. We also present an alternative assembly including 27 Gb raw reads generated using the Pacific Biosciences platform. In addition, we sequenced the proteome of the same individual and RNA from 3 different tissue types from 3 other species of squid (Onychoteuthis banksii, Dosidicus gigas, and Sthenoteuthis oualaniensis) to assist genome annotation. We annotated 33,406 protein-coding genes supported by evidence, and the genome completeness estimated by BUSCO reached 92%. Repetitive regions cover 49.17% of the genome. CONCLUSIONS: This annotated draft genome of A. dux provides a critical resource to investigate the unique traits of this species, including its gigantism and key adaptations to deep-sea environments.

DeepIso: A Deep Learning Model for Peptide Feature Detection from LC-MS map.

Liquid chromatography with tandem mass spectrometry (LC-MS/MS) based quantitative proteomics provides the relative different protein abundance in healthy and disease-afflicted patients, which offers the information for molecular interactions, signaling pathways, and biomarker identification to serve the drug discovery and clinical research. Typical analysis workflow begins with the peptide feature detection and intensity calculation from LC-MS map. We are the first to propose a deep learning based model, DeepIso, that combines recent advances in Convolutional Neural Network (CNN) and Recurrent Neural Network (RNN) to detect peptide features of different charge states, as well as, estimate their intensity. Existing tools are designed with limited engineered features and domain-specific parameters, which are hardly updated despite a huge amount of new coming proteomic data. On the other hand, DeepIso consisting of two separate deep learning based modules, learns multiple levels of representation of high dimensional data itself through many layers of neurons, and adaptable to newly acquired data. The peptide feature list reported by our model matches with 97.43% of high quality MS/MS identifications in a benchmark dataset, which is higher than the matching produced by several widely used tools. Our results demonstrate that novel deep learning tools are desirable to advance the state-of-the-art in protein identification and quantification.

Hospital-based zoonotic disease surveillance in Bangladesh: design, field data and difficulties.

Early detection of zoonotic diseases allows for the implementation of early response measures, reducing loss of human life and economic disruption. We implemented a surveillance system in hospitals in Bangladesh to screen acutely ill hospitalized patients with severe respiratory infection and meningoencephalitis for zoonotic exposures. Patients were screened for the risk of zoonotic exposures with five questions covering vocational exposures, sick domestic animal and wild animal contact, and date palm sap consumption in the three weeks preceding illness onset. Patients giving at least one positive response were considered a potential zoonotic exposure. From September 2013 to March 2017, a total of 11 429 hospitalized patients across 14 participating hospitals were screened for exposures. Overall, 2% of patients reported a potential zoonotic exposure in the three-week period prior to becoming ill. Sixteen per cent of hospitalized patients with reported exposures died. After routine surveillance diagnostic testing, 88% of patients admitted to the hospital after a potential zoonotic exposure did not have a laboratory diagnosed aetiology for their illness. Hospital-based surveillance systems such as the Bangladeshi example presented here could play an important future role in the early detection of zoonotic spillover diseases. This article is part of the theme issue 'Dynamic and integrative approaches to understanding pathogen spillover'.

Complete De Novo Assembly of Monoclonal Antibody Sequences.

De novo protein sequencing is one of the key problems in mass spectrometry-based proteomics, especially for novel proteins such as monoclonal antibodies for which genome information is often limited or not available. However, due to limitations in peptides fragmentation and coverage, as well as ambiguities in spectra interpretation, complete de novo assembly of unknown protein sequences still remains challenging. To address this problem, we propose an integrated system, ALPS, which for the first time can automatically assemble full-length monoclonal antibody sequences. Our system integrates de novo sequencing peptides, their quality scores and error-correction information from databases into a weighted de Bruijn graph to assemble protein sequences. We evaluated ALPS performance on two antibody data sets, each including a heavy chain and a light chain. The results show that ALPS was able to assemble three complete monoclonal antibody sequences of length 216-441 AA, at 100% coverage, and 96.64-100% accuracy.

Dynamics of Japanese encephalitis virus transmission among pigs in Northwest Bangladesh and the potential impact of pig vaccination.

BACKGROUND: Japanese encephalitis (JE) virus infection can cause severe disease in humans, resulting in death or permanent neurologic deficits among survivors. Studies indicate that the incidence of JE is high in northwestern Bangladesh. Pigs are amplifying hosts for JE virus (JEV) and a potentially important source of virus in the environment. The objectives of this study were to describe the transmission dynamics of JEV among pigs in northwestern Bangladesh and estimate the potential impact of vaccination to reduce incidence among pigs. METHODOLOGY/PRINCIPAL FINDINGS: We conducted a comprehensive census of pigs in three JE endemic districts and tested a sample of them for evidence of previous JEV infection. We built a compartmental model to describe JEV transmission dynamics in this region and to estimate the potential impact of pig vaccination. We identified 11,364 pigs in the study area. Previous JEV infection was identified in 30% of pigs with no spatial differences in the proportion of pigs that were seropositive across the study area. We estimated that JEV infects 20% of susceptible pigs each year and the basic reproductive number among pigs was 1.2. The model suggest that vaccinating 50% of pigs each year resulted in an estimated 82% reduction in annual incidence in pigs. CONCLUSIONS/SIGNIFICANCE: The widespread distribution of historic JEV infection in pigs suggests they may play an important role in virus transmission in this area. Future studies are required to understand the contribution of pig infections to JE risk in humans and the potential impact of pig vaccination on human disease.

Hydroclimatological variability and dengue transmission in Dhaka, Bangladesh: a time-series study.

BACKGROUND: While floods can potentially increase the transmission of dengue, only few studies have reported the association of dengue epidemics with flooding. We estimated the effects of river levels and rainfall on the hospital admissions for dengue fever at 11 major hospitals in Dhaka, Bangladesh. METHODS: We examined time-series of the number of hospital admissions of dengue fever in relation to river levels from 2005 to 2009 using generalized linear Poisson regression models adjusting for seasonal, between-year variation, public holidays and temperature. RESULTS: There was strong evidence for an increase in dengue fever at high river levels. Hospitalisations increased by 6.9% (95% CI: 3.2, 10.7) for each 0.1 metre increase above a threshold (3.9 metres) for the average river level over lags of 0-5 weeks. Conversely, the number of hospitalisations increased by 29.6% (95% CI: 19.8, 40.2) for a 0.1 metre decrease below the same threshold of the average river level over lags of 0-19 weeks. CONCLUSIONS: Our findings provide evidence that factors associated with both high and low river levels increase the hospitalisations of dengue fever cases in Dhaka.

Pandemic strains of O3:K6 Vibrio parahaemolyticus in the aquatic environment of Bangladesh.

A total of 1500 environmental strains of Vibrio parahaemolyticus, isolated from the aquatic environment of Bangladesh, were screened for the presence of a major V. parahaemolyticus virulence factor, the thermostable direct haemolysin (tdh) gene, by the colony blot hybridization method using a digoxigenin-labeled tdh gene probe. Of 1500 strains, 5 carried the tdh sequence, which was further confirmed by PCR using primers specific for the tdh gene. Examination by PCR confirmed that the 5 strains were V. parahaemolyticus and lacked the thermostable direct haemolysin-related haemolysin (trh) gene, the alternative major virulence gene known to be absent in pandemic strains. All 5 strains gave positive Kanagawa phenomenon reaction with characteristic beta-haemolysis on Wagatsuma agar medium. Southern blot analysis of the HindIII-digested chromosomal DNA demonstrated, in all 5 strains, the presence of 2 tdh genes common to strains positive for Kanagawa phenomenon. However, the 5 strains were found to belong to 3 different serotypes (O3:K29, O4:K37, and O3:K6). The 2 with pandemic serotype O3:K6 gave positive results in group-specific PCR and ORF8 PCR assays, characteristics unique to the pandemic clone. Clonal variations among the 5 isolates were analyzed by comparing RAPD and ribotyping patterns. Results showed different patterns for the 3 serotypes, but the pattern was identical among the O3:K6 strains. This is the first report on the isolation of pandemic O3:K6 strains of V. parahaemolyticus from the aquatic environment of Bangladesh.