Browse
Submit Data
Databases
API
Help

Dataset Information

6 Views

0 Connections

0 Citations

0 Reanalyses

0 Downloads

Omics score: 0

Preparing for the next pandemic via transfer learning from existing diseases with hierarchical multi-modal BERT: a study on COVID-19 outcome prediction.

ABSTRACT: Developing prediction models for emerging infectious diseases from relatively small numbers of cases is a critical need for improving pandemic preparedness. Using COVID-19 as an exemplar, we propose a transfer learning methodology for developing predictive models from multi-modal electronic healthcare records by leveraging information from more prevalent diseases with shared clinical characteristics. Our novel hierarchical, multi-modal model ([Formula: see text]) integrates baseline risk factors from the natural language processing of clinical notes at admission, time-series measurements of biomarkers obtained from laboratory tests, and discrete diagnostic, procedure and drug codes. We demonstrate the alignment of [Formula: see text]'s predictions with well-established clinical knowledge about COVID-19 through univariate and multivariate risk factor driven sub-cohort analysis. [Formula: see text]'s superior performance over state-of-the-art methods shows that leveraging patient data across modalities and transferring prior knowledge from similar disorders is critical for accurate prediction of patient outcomes, and this approach may serve as an important tool in the early response to future pandemics.

SUBMITTER: Agarwal K

PROVIDER: S-EPMC9232529 | biostudies-literature |

REPOSITORIES: biostudies-literature

ACCESS DATA

Json Xml

Similar Datasets

BERT-TFBS: a novel BERT-based model for predicting transcription factor binding sites by transfer learning.

Project description:Transcription factors (TFs) are proteins essential for regulating genetic transcriptions by binding to transcription factor binding sites (TFBSs) in DNA sequences. Accurate predictions of TFBSs can contribute to the design and construction of metabolic regulatory systems based on TFs. Although various deep-learning algorithms have been developed for predicting TFBSs, the prediction performance needs to be improved. This paper proposes a bidirectional encoder representations from transformers (BERT)-based model, called BERT-TFBS, to predict TFBSs solely based on DNA sequences. The model consists of a pre-trained BERT module (DNABERT-2), a convolutional neural network (CNN) module, a convolutional block attention module (CBAM) and an output module. The BERT-TFBS model utilizes the pre-trained DNABERT-2 module to acquire the complex long-term dependencies in DNA sequences through a transfer learning approach, and applies the CNN module and the CBAM to extract high-order local features. The proposed model is trained and tested based on 165 ENCODE ChIP-seq datasets. We conducted experiments with model variants, cross-cell-line validations and comparisons with other models. The experimental results demonstrate the effectiveness and generalization capability of BERT-TFBS in predicting TFBSs, and they show that the proposed model outperforms other deep-learning models. The source code for BERT-TFBS is available at https://github.com/ZX1998-12/BERT-TFBS.

| S-EPMC11066948 | biostudies-literature

Preparing for the Next Wave of COVID-19: Resilience in the Face of a Spreading Pandemic.

Project description:COVID-19 painfully demonstrates how little resilience our societies have to novel viruses. Societies, decision makers, and scientists lack (1) a comprehensive understanding of the complexity of viral outbreaks and their impact on society; (2) intervention portfolios; and (3) a global crisis and resilience policy, all of which are required to develop appropriate measures and to improve societal resilience. We highlight COVID-19 immunity as one key benchmark in preparation for the next wave of the pandemic. Specifically, using network scenarios, we demonstrate the substantial advantage of reintegrating health care workers with acquired COVID-19 immunity in epidemic hotspots, which would not only enable their safe contribution to the health care system but also drastically contain further spread.

| S-EPMC7312045 | biostudies-literature

Preparing for the Next Normal: Transformation in the Role of Medical Affairs Following the COVID-19 Pandemic.

Project description:The medical affairs function represents one of the scientific interfaces in a pharmaceutical organization. Over the last two decades, medical affairs has evolved from being a support function to a strategic pillar within organizational business units. The COVID-19 pandemic has given rise to unforeseen circumstances resulting in a dramatic change in external stakeholder engagements, catapulting the medical affairs function into leading the way on scientific engagements and patient-centric endeavors. The changes in stakeholder interactions and behavior as a result of the pandemic last year are likely to persist in the foreseeable future for which medical affairs professionals need to enhance existing skill sets and acquire expertise in newer domains. In this paper, the transformation of the medical affairs team to a key strategic partner and the skills required to strengthen this transition, in the next normal of a post-COVID world, is explored.

| S-EPMC8256774 | biostudies-literature

Accurate TCR-pMHC interaction prediction using a BERT-based transfer learning method.

Project description:Accurate prediction of TCR-pMHC binding is important for the development of cancer immunotherapies, especially TCR-based agents. Existing algorithms often experience diminished performance when dealing with unseen epitopes, primarily due to the complexity in TCR-pMHC recognition patterns and the scarcity of available data for training. We have developed a novel deep learning model, 'TCR Antigen Binding Recognition' based on BERT, named as TABR-BERT. Leveraging BERT's potent representation learning capabilities, TABR-BERT effectively captures essential information regarding TCR-pMHC interactions from TCR sequences, antigen epitope sequences and epitope-MHC binding. By transferring this knowledge to predict TCR-pMHC recognition, TABR-BERT demonstrated better results in benchmark tests than existing methods, particularly for unseen epitopes.

| S-EPMC10783865 | biostudies-literature

Preparing for the next viral threat with broad-spectrum antivirals.

Project description:There is a large global unmet need for the development of countermeasures to combat hundreds of viruses known to cause human disease and for the establishment of a therapeutic portfolio for future pandemic preparedness. Most approved antiviral therapeutics target proteins encoded by a single virus, providing a narrow spectrum of coverage. This, combined with the slow pace and high cost of drug development, limits the scalability of this direct-acting antiviral (DAA) approach. Here, we summarize progress and challenges in the development of broad-spectrum antivirals that target either viral elements (proteins, genome structures, and lipid envelopes) or cellular proviral factors co-opted by multiple viruses via newly discovered compounds or repurposing of approved drugs. These strategies offer new means for developing therapeutics against both existing and emerging viral threats that complement DAAs.

| S-EPMC10232003 | biostudies-literature

Towards the future exploration of mucosal mRNA vaccines against emerging viral diseases; lessons from existing next-generation mucosal vaccine strategies.

Project description:The mRNA vaccine platform has offered the greatest potential in fighting the COVID-19 pandemic owing to rapid development, effectiveness, and scalability to meet the global demand. There are many other mRNA vaccines currently being developed against different emerging viral diseases. As with the current COVID-19 vaccines, these mRNA-based vaccine candidates are being developed for parenteral administration via injections. However, most of the emerging viruses colonize the mucosal surfaces prior to systemic infection making it very crucial to target mucosal immunity. Although parenterally administered vaccines would induce a robust systemic immunity, they often provoke a weak mucosal immunity which may not be effective in preventing mucosal infection. In contrast, mucosal administration potentially offers the dual benefit of inducing potent mucosal and systemic immunity which would be more effective in offering protection against mucosal viral infection. There are however many challenges posed by the mucosal environment which impede successful mucosal vaccination. The development of an effective delivery system remains a major challenge to the successful exploitation of mucosal mRNA vaccination. Nonetheless, a number of delivery vehicles have been experimentally harnessed with different degrees of success in the mucosal delivery of mRNA vaccines. In this review, we provide a comprehensive overview of mRNA vaccines and summarise their application in the fight against emerging viral diseases with particular emphasis on COVID-19 mRNA platforms. Furthermore, we discuss the prospects and challenges of mucosal administration of mRNA-based vaccines, and we explore the existing experimental studies on mucosal mRNA vaccine delivery.

| S-EPMC9239993 | biostudies-literature

Lack of impact of pre-existing T97A HIV-1 integrase mutation on integrase strand transfer inhibitor resistance and treatment outcome.

Project description:T97A is an HIV-1 integrase polymorphism associated with integrase strand transfer inhibitor (INSTI) resistance. Using pooled data from 16 clinical studies, we investigated the prevalence of T97A (pre-existing and emergent) and its impact on INSTI susceptibility and treatment response in INSTI-naive patients who enrolled on elvitegravir (EVG)- or raltegravir (RAL)-based regimens. Prior to INSTI-based therapy, primary INSTI resistance-associated mutations (RAMs) were absent and T97A pre-existed infrequently (1.4%; 47 of 3367 integrase sequences); most often among non-B (5.3%) than B (0.9%) HIV-1 subtypes. During INSTI-based therapy, few patients experienced virologic failure with emergent INSTI RAMs (3%; 122 of 3881 patients), among whom T97A emerged infrequently in the presence (n = 6) or absence (n = 8) of primary INSTI RAMs. A comparison between pre-existing and emergent T97A patient populations (i.e., in the absence of primary INSTI RAMs) showed no significant differences in EVG or RAL susceptibility in vitro. Furthermore, among all T97A-containing viruses tested, only 38-44% exhibited reduced susceptibility to EVG and/or RAL (all of low magnitude; <11-fold), while all maintained susceptibility to dolutegravir. Of the patients with pre-existing T97A, 17 had available clinical follow-up: 16 achieved virologic suppression and 1 maintained T97A and INSTI sensitivity without further resistance development. Overall, T97A is an infrequent integrase polymorphism that is enriched among non-B HIV-1 subtypes and can confer low-level reduced susceptibility to EVG and/or RAL. However, detection of T97A does not affect response to INSTI-based therapy with EVG or RAL. These results suggest a very low risk of initiating INSTI-based therapy in patients with pre-existing T97A.

| S-EPMC5315389 | biostudies-literature

The role of clinical simulation in preparing for a pandemic.

Project description: Not available

| S-EPMC7836946 | biostudies-literature

Cross-modal transfer in visual and nonvisual cues in bumblebees.

Project description:Bumblebees Bombus terrestris are good at learning to distinguish between patterned flowers. They can differentiate between flowers that differ only in their patterning of scent, surface texture, temperature, or electrostatic charge, in addition to visual patterns. As recently shown, bumblebees trained to discriminate between nonvisual scent patterns can transfer this learning to visually patterned flowers that show similar spatial patterning to the learnt scent patterns. Bumblebees can, therefore, transfer learnt patterns between different sensory modalities, without needing to relearn them. We used differential conditioning techniques to explore whether cross-modal transfer of learnt patterns also occurred between visual and temperature patterns. Bumblebees that successfully learnt to distinguish rewarding and unrewarding temperature patterns did not show any preferences for the corresponding unlearnt visual pattern. Similarly, bumblebees that learnt visual patterns did not transfer these to temperature patterns, suggesting that they are unable to transfer learning of temperature and visual patterns. We discuss how cross-modality pattern learning may be limited to modalities that have potentially strong neurological links, such as the previously demonstrated transfer between scent and visual patterns.

| S-EPMC6579774 | biostudies-literature

A hierarchical Bayesian model for next-generation population genomics.

Project description:The demography of populations and natural selection shape genetic variation across the genome and understanding the genomic consequences of these evolutionary processes is a fundamental aim of population genetics. We have developed a hierarchical Bayesian model to quantify genome-wide population structure and identify candidate genetic regions affected by selection. This model improves on existing methods by accounting for stochastic sampling of sequences inherent in next-generation sequencing (with pooled or indexed individual samples) and by incorporating genetic distances among haplotypes in measures of genetic differentiation. Using simulations we demonstrate that this model has a low false-positive rate for classifying neutral genetic regions as selected genes (i.e., ?(ST) outliers), but can detect recent selective sweeps, particularly when genetic regions in multiple populations are affected by selection. Nonetheless, selection affecting just a single population was difficult to detect and resulted in a high false-negative rate under certain conditions. We applied the Bayesian model to two large sets of human population genetic data. We found evidence of widespread positive and balancing selection among worldwide human populations, including many genetic regions previously thought to be under selection. Additionally, we identified novel candidate genes for selection, several of which have been linked to human diseases. This model will facilitate the population genetic analysis of a wide range of organisms on the basis of next-generation sequence data.

| S-EPMC3063681 | biostudies-literature

OmicsDI is part of the ELIXIR infrastructure

OmicsDI is an Elixir interoperability service. Learn more ›

Tweets

OmicsDI Databases

PRIDE
PeptideAtlas
MassIVE
JPOST Repository
Physiome Model Repository

EGA
EVA
ENA
LINCS
PAXDB
Cell Collective

MetaboLights
Metabolomics Workbench
MetabolomeExpress
GNPS
BioModels
FAIRDOMHub

ArrayExpress
dbGaP
ExpressionAtlas
GEO
NODE

Information

Databases
Help
API
Contact us
Code on GitHub
Terms of use
Submit Data