Category: Statistics

J Comput Biol. 2023 Nov;30(11):1226-1239. doi: 10.1089/cmb.2023.0208.

ABSTRACT

Statistical models that accurately predict the binding affinity of an input ligand-protein pair can greatly accelerate drug discovery. Such models are trained on available ligand-protein interaction data sets, which may contain biases that lead the predictor models to learn data set-specific, spurious patterns instead of generalizable relationships. This leads the prediction performances of these models to drop dramatically for previously unseen biomolecules. Various approaches that aim to improve model generalizability either have limited applicability or introduce the risk of degrading overall prediction performance. In this article, we present DebiasedDTA, a novel training framework for drug-target affinity (DTA) prediction models that addresses data set biases to improve the generalizability of such models. DebiasedDTA relies on reweighting the training samples to achieve robust generalization, and is thus applicable to most DTA prediction models. Extensive experiments with different biomolecule representations, model architectures, and data sets demonstrate that DebiasedDTA achieves improved generalizability in predicting drug-target affinities.

PMID:37988395 | DOI:10.1089/cmb.2023.0208

PLoS One. 2023 Nov 21;18(11):e0294445. doi: 10.1371/journal.pone.0294445. eCollection 2023.

ABSTRACT

This paper conducts a systematic statistical analysis of the characteristics of the geographical empirical distributions for the numbers of both cumulative and daily confirmed COVID-19 cases and deaths at county, city, and state levels over a time span from January 2020 to June 2022. The mathematical heavy-tailed distributions can be used for fitting the empirical distributions observed in different temporal stages and geographical scales. The estimations of the shape parameter of the tail distributions using the Generalized Pareto Distribution also support the observations of the heavy-tailed distributions. According to the characteristics of the heavy-tailed distributions, the evolution course of the geographical empirical distributions can be divided into three distinct phases, namely the power-law phase, the lognormal phase I, and the lognormal phase II. These three phases could serve as an indicator of the severity degree of the COVID-19 pandemic within an area. The empirical results suggest important intrinsic dynamics of a human infectious virus spread in the human interconnected physical complex network. The findings extend previous empirical studies and could provide more strict constraints for current mathematical and physical modeling studies, such as the SIR model and its variants based on the theory of complex networks.

PMID:37988387 | DOI:10.1371/journal.pone.0294445

PLOS Glob Public Health. 2023 Nov 21;3(11):e0001990. doi: 10.1371/journal.pgph.0001990. eCollection 2023.

ABSTRACT

Bacterial sepsis is generally a major concern in ill infants. To help triaging decisions by front-line health workers in these situations, the World Health Organization (WHO) has developed danger signs (DS). The objective of this study was to evaluate the extent to which nine DS predict bacterial sepsis in young infants presenting with suspected sepsis in a low-income country setting. The study pragmatically evaluated nine DS in infants younger than 3 months with suspected sepsis in a regional hospital in Lilongwe, Malawi, between June 2018 and April 2020. Main outcomes were positive blood or cerebrospinal fluid (CSF) cultures for neonatal pathogens, and mortality. Among 401 infants (gestational age [mean ± SD]: 37.1±3.3 weeks, birth weight 2865±785 grams), 41 had positive blood or CSF cultures for a neonatal pathogen. In-hospital mortality occurred in 9.7% of infants overall (N = 39/401), of which 61.5% (24/39) occurred within 48 hours of admission. Mortality was higher in infants with bacterial sepsis compared to other infants (22.0% [9/41] versus 8.3% [30/360]; p = 0.005). All DS were associated with mortality except for temperature instability and tachypnea, whereas none of the DS were significantly associated with bacterial sepsis, except for “unable to feed” (OR 2.25; 95%CI: 1.17-4.44; p = 0.017). The number of DS predicted mortality (OR: 1.75; 95%CI: 1.43-2.17; p<0.001; AUC: 0.756), but was marginally associated with positive cultures with a neonatal pathogen (OR 1.22; 95%CI: 1.00-1.49; p = 0.046; AUC: 0.743). The association between number of DS and mortality remained significant after adjusting for admission weight, the only statistically significant co-variable (OR 1.75 [95% CI: 1.39-2.23]; p<0.001). Considering all positive cultures including potential bacterial contaminants resulted a non-significant association between number of DS and sepsis (OR 1.09 [95% CI: 0.93-1.28]; p = 0.273). In conclusion, this study shows that DS were strongly associated with death, but were marginally associated with culture-positive pathogen sepsis in a regional hospital setting. These data imply that the incidence of bacterial sepsis and attributable mortality in infants in LMIC settings may be inaccurately estimated based on clinical signs alone.

PMID:37988384 | DOI:10.1371/journal.pgph.0001990

J Acoust Soc Am. 2023 Nov 1;154(5):3374-3387. doi: 10.1121/10.0022459.

ABSTRACT

Target strength (TS) is commonly used to convert acoustic backscatter from marine organisms to numerical abundance estimates. Shipboard, tank-based TS measurements were made on four sub-Arctic krill species (Euphausia pacifica, Thysanoessa spinifera, Thysanoessa inermis, and Thysanoessa raschii) from the eastern Bering Sea and Gulf of Alaska at discrete frequencies between 42 and 455 kHz. These measurements were compared to scattering model predictions parameterized with data from the same (when possible) individual krill. Statistically significant differences between modeled and experimental estimates at 42, 45, 120, and 131 kHz exceeded 2 dB on average. Variability in the signal-to-noise ratio, animal length, and measurements from two separate narrowband and broadband transducer pairs (at those frequencies) did not account for these differences. Scattering predictions at 120 and 131 kHz were consistent with the expected transition from Rayleigh-to-geometric scattering where models become increasingly sensitive to orientation and body shape variability. Disagreement between modeled and measured TS may be due to using scattering models developed for, and validated on, larger krill (i.e., Euphausia superba) rather than smaller species of krill. Acoustic surveys of smaller (15-30 mm) krill may require further validation of both the generalizability and parameterization of applied scattering models.

PMID:37988372 | DOI:10.1121/10.0022459

Neuropsychol Rehabil. 2023 Dec;33(10):1650-1671. doi: 10.1080/09602011.2022.2140680. Epub 2022 Nov 3.

ABSTRACT

Emotional disorders are pervasive in the acquired brain injury (ABI) population, adversely affecting quality of life and rehabilitation. This study aimed to explore the unique associative effects of alexithymia as measured by the Perth Alexithymia Questionnaire (PAQ; i.e., difficulty identifying positive/negative feelings, difficulty describing positive/negative feelings, and externally orientated thinking), on emotional outcomes as measured by the Depression Anxiety Stress Scale-21 (DASS-21) and Mayo-Portland Adaptability Inventory (MPAI-4) Adjustment index, in 83 adults with ABI. The addition of alexithymia to hierarchical multiple regression models (controlling for demographic, injury-related, and functional outcome variables) yielded statistically significant changes in R² for all emotional outcome measures (i.e., Depression, Anxiety, Stress, and Adjustment). Difficulty identifying negative feelings was found to be a significant unique predictor of Depression (β = .43 p = <.001), Anxiety (β = .40, p <.001), Stress (β = .49, p <.001), and Adjustment (β = .26, p = .001). Externally oriented thinking was found to be a significant unique predictor of Adjustment (β = -.15, p = .033). These findings strengthen the argument that alexithymia, especially difficulties identifying negative feelings, may be an important risk factor for psychological distress in ABI and should be considered during early rehabilitation.

PMID:37988367 | DOI:10.1080/09602011.2022.2140680

PLoS One. 2023 Nov 21;18(11):e0291348. doi: 10.1371/journal.pone.0291348. eCollection 2023.

ABSTRACT

The impact of the COVID-19 pandemic on University students has been a topic of fiery debate and of public health research. This study demonstrates the use of a combination of spatiotemporal epidemiological models to describe the trends in COVID-19 positive cases on spatial, temporal and spatiotemporal scales. In addition, this study proposes new epidemiological metrics to describe the connectivity between observed positivity; an analogous metric to the R number in conventional epidemiology. The proposed indices, Rspatial, Rspatiotemporal and Rscaling will aim to improve the characterisation of the spread of infectious disease beyond that of the COVID-19 framework and as a result inform relevant public health policy. Apart from demonstrating the application of the novel epidemiological indices, the key findings in this study are: firstly, there were some Intermediate Zones in Edinburgh with noticeably high levels of COVID-19 positivity, and that the first outbreak during the study period was observed in Dalry and Fountainbridge. Secondly, the estimation of the distance over which the COVID-19 counts at the halls of residence are spatially correlated (or related to each other) was found to be 0.19km (0.13km to 0.27km) and is denoted by the index, Rspatial. This estimate is useful for public health policy in this setting, especially with contact tracing. Thirdly, the study indicates that the association between the surrounding community level of COVID-19 positivity (Intermediate Zones in Edinburgh) and that of the University of Edinburgh’s halls of residence was not statistically significant. Fourthly, this study reveals that relatively high levels of COVID-19 positivity were observed for halls for which higher COVID-19 fines were issued (Spearman’s correlation coefficient = 0.34), and separately, for halls which were non-ensuite relatively to those which were not (Spearman’s correlation coefficient = 0.16). Finally, Intermediate Zones with the highest positivity were associated with student residences that experienced relatively high COVID-19 positivity (Spearman’s correlation coefficient = 0.27).

PMID:37988358 | DOI:10.1371/journal.pone.0291348

JMIR Dermatol. 2023 Nov 21;6:e49950. doi: 10.2196/49950.

ABSTRACT

BACKGROUND: Oculocutaneous albinism is a congenital disorder that causes hypopigmentation of the skin, hair, and eyes due to a lack of melanin. People with albinism are at increased risk of developing skin complications, such as solar keratosis and skin cancers, leading to higher morbidity. As education is crucial in managing albinism, leveraging information technology, such as WhatsApp, can provide an effective intervention for digital health education.

OBJECTIVE: This study aims to assess the impact of WhatsApp as a tool for providing health education among people with albinism.

METHODS: The design of the study was interventional. The intervention consisted of weekly health education sessions conducted in a WhatsApp group for the duration of 4 weeks. The topics discussed were knowledge of albinism, sun protection practices, the use of sunscreen, and myths about albinism. They were all covered in 4 WhatsApp sessions held in 4 separate days. A web-based questionnaire was filled out before and after the intervention by the participants. Mann-Whitney U test was used to compare the pre- and postknowledge scores. Spearman correlation was used to correlate data.

RESULTS: The mean age of the study participants was 28.28 (SD 11.57) years. The number of participants was 140 in the preintervention period and 66 in the postintervention period. A statistically significant increase in overall knowledge (P=.01), knowledge of sunscreen (P=.01), and knowledge of sun protection (P<.01) was observed following the intervention. Before the intervention, a positive correlation was observed between age (r=0.17; P=.03) and education level (r=0.19; P=.02) with participants’ overall knowledge. However, after the intervention, there was no significant correlation between knowledge and age or education level. A percentage increase of 5.23% was observed in the overall knowledge scores following the intervention.

CONCLUSIONS: WhatsApp is an effective tool for educating people with albinism and can act as an alternative to the conventional methods of health education. It shows promising outcomes irrespective of the health literacy level of people with albinism. This educational intervention can positively impact behavior change and translate to consistent sun protection practices. The limitations of this study include the possibility of social desirability bias and data security.

PMID:37988154 | DOI:10.2196/49950

Bioinformatics. 2023 Nov 21:btad703. doi: 10.1093/bioinformatics/btad703. Online ahead of print.

ABSTRACT

SUMMARY: Federated learning enables collaboration in medicine, where data is scattered across multiple centers without the need to aggregate the data in a central cloud. While, in general, machine learning models can be applied to a wide range of data types, graph neural networks (GNNs) are particularly developed for graphs, which are very common in the biomedical domain. For instance, a patient can be represented by a protein-protein interaction (PPI) network where the nodes contain the patient-specific omics features. Here, we present our Ensemble-GNN software package, which can be used to deploy federated, ensemble-based GNNs in Python. Ensemble-GNN allows to quickly build predictive models utilizing PPI networks consisting of various node features such as gene expression and/or DNA methylation. We exemplary show the results from a public dataset of 981 patients and 8469 genes from the Cancer Genome Atlas (TCGA).

AVAILABILITY: The source code is available at https://github.com/pievos101/Ensemble-GNN, and the data at Zenodo (DOI: 10.5281/zenodo.8305122).

SUPPLEMENTARY INFORMATION: Supplementary data are available at Bioinformatics online.

PMID:37988152 | DOI:10.1093/bioinformatics/btad703

JMIR Mhealth Uhealth. 2023 Nov 21;11:e49144. doi: 10.2196/49144.

ABSTRACT

BACKGROUND: Patient-generated health data are important in the management of several diseases. Although there are limitations, information can be obtained using a wearable device and time-related information such as exercise time or sleep time can also be obtained. Fitbits can be used to acquire sleep onset, sleep offset, total sleep time (TST), and wakefulness after sleep onset (WASO) data, although there are limitations regarding the depth of sleep and satisfaction; therefore, the patient’s subjective response is still important information that cannot be replaced by wearable devices.

OBJECTIVE: To effectively use patient-generated health data related to time such as sleep, it is first necessary to understand the characteristics of the time response recorded by the user. Therefore, the aim of this study was to analyze the characteristics of individuals’ time perception in comparison with wearable data.

METHODS: Sleep data were acquired for 2 weeks using a Fitbit. Participants’ sleep records were collected daily through chatbot conversations while wearing the Fitbit, and the two sets of data were statistically compared.

RESULTS: In total, 736 people aged 30-59 years were recruited for this study, and the sleep data of 543 people who wore a Fitbit and responded to the chatbot for more than 7 days on the same day were analyzed. Research participants tended to respond to sleep-related times on the hour or in 30-minute increments, and each participant responded within the range of 60-90 minutes from the value measured by the Fitbit. On average for all participants, the chat responses and the Fitbit data were similar within a difference of approximately 15 minutes. Regarding sleep onset, the participant response was 8 minutes and 39 seconds (SD 58 minutes) later than that of the Fitbit data, whereas with respect to sleep offset, the response was 5 minutes and 38 seconds (SD 57 minutes) earlier. The participants’ actual sleep time (AST) indicated in the chat was similar to that obtained by subtracting the WASO from the TST measured by the Fitbit. The AST was 13 minutes and 39 seconds (SD 87 minutes) longer than the time WASO was subtracted from the Fitbit TST. On days when the participants reported good sleep, they responded 19 (SD 90) minutes longer on the AST than the Fitbit data. However, for each sleep event, the probability that the participant’s AST was within ±30 and ±60 minutes of the Fitbit TST-WASO was 50.7% and 74.3%, respectively.

CONCLUSIONS: The chatbot sleep response and Fitbit measured time were similar on average and the study participants had a slight tendency to perceive a relatively long sleep time if the quality of sleep was self-reported as good. However, on a participant-by-participant basis, it was difficult to predict participants’ sleep duration responses with Fitbit data. Individual variations in sleep time perception significantly affect patient responses related to sleep, revealing the limitations of objective measures obtained through wearable devices.

PMID:37988148 | DOI:10.2196/49144

QJM. 2023 Nov 21:hcad262. doi: 10.1093/qjmed/hcad262. Online ahead of print.

ABSTRACT

BACKGROUND: Olfactory and gustatory dysfunctions (OGDs) are key symptoms of COVID-19, which may lead to neurological complications, and lack of effective treatment. This may be because post-disease treatments may be too late to protect the olfactory and gustatory functions.

AIM: To evaluate the effectiveness of early use of saline nasal irrigation (SNI), corticosteroid nasal spray, and saline or chlorhexidine gluconate mouthwash for preventing OGDs in COVID-19.

DESIGN: This study was a double-blind randomized controlled trial.

METHODS: The study was conducted from May 5 to June 16, 2022. We recruited patients from three hospitals who were admitted with COVID-19 but without OGDs on the day of admission. Olfactory and gustatory functions were evaluated using the Taste and Smell Survey and the numerical visual analog scale. Participants were randomized to the saline, drug, or control groups. The control group received no intervention, saline group received SNI plus saline nasal spray and mouthwash, and the trial group received SNI plus budesonide nasal spray and chlorhexidine gluconate mouthwash. Participants were assessed again on the day of discharge.

RESULTS: A total of 379 patients completed the trial. The prevalence of OGDs was significantly lower in the saline (11.8%, 95% CI, 6.6-19.0%; P < 0.001) and trial (8.3%, 95% CI, 4.1-14.8%; P < 0.001) groups than in the control group (40.0%, 95% CI, 31.8-48.6%). Additionally, both interventions reduced the severity of OGDs.

CONCLUSIONS: We demonstrated effective strategies for preventing COVID-19-related OGDs, and the findings may guide early management of SARS-CoV-2 infection to reduce the incidence of COVID-19-related complications.

PMID:37988146 | DOI:10.1093/qjmed/hcad262