Faecal microbiome-based machine learning for multi-class disease diagnosis.

Systemic characterisation of the human faecal microbiome provides the opportunity to develop non-invasive approaches in the diagnosis of a major human disease. However, shared microbial signatures across different diseases make accurate diagnosis challenging in single-disease models. Herein, we present a machine-learning multi-class model using faecal metagenomic dataset of 2,320 individuals with nine well-characterised phenotypes, including colorectal cancer, colorectal adenomas, Crohn's disease, ulcerative colitis, irritable bowel syndrome, obesity, cardiovascular disease, post-acute COVID-19 syndrome and healthy individuals. Our processed data covers 325 microbial species derived from 14.3 terabytes of sequence. The trained model achieves an area under the receiver operating characteristic curve (AUROC) of 0.90 to 0.99 (Interquartile range, IQR, 0.91-0.94) in predicting different diseases in the independent test set, with a sensitivity of 0.81 to 0.95 (IQR, 0.87-0.93) at a specificity of 0.76 to 0.98 (IQR 0.83-0.95). Metagenomic analysis from public datasets of 1,597 samples across different populations observes comparable predictions with AUROC of 0.69 to 0.91 (IQR 0.79-0.87). Correlation of the top 50 microbial species with disease phenotypes identifies 363 significant associations (FDR < 0.05). This microbiome-based multi-disease model has potential clinical application in disease diagnostics and treatment response monitoring and warrants further exploration.

Reciprocal perspective as a super learner improves drug-target interaction prediction (MUSDTI).

The identification of novel drug-target interactions (DTI) is critical to drug discovery and drug repurposing to address contemporary medical and public health challenges presented by emergent diseases. Historically, computational methods have framed DTI prediction as a binary classification problem (indicating whether or not a drug physically interacts with a given protein target); however, framing the problem instead as a regression-based prediction of the physiochemical binding affinity is more meaningful. With growing databases of experimentally derived drug-target interactions (e.g. Davis, Binding-DB, and Kiba), deep learning-based DTI predictors can be effectively leveraged to achieve state-of-the-art (SOTA) performance. In this work, we formulated a DTI competition as part of the coursework for a senior undergraduate machine learning course and challenged students to generate component DTI models that might surpass SOTA models and effectively combine these component models as part of a meta-model using the Reciprocal Perspective (RP) multi-view learning framework. Following 6 weeks of concerted effort, 28 student-produced component deep-learning DTI models were leveraged in this work to produce a new SOTA RP-DTI model, denoted the Meta Undergraduate Student DTI (MUSDTI) model. Through a series of experiments we demonstrate that (1) RP can considerably improve SOTA DTI prediction, (2) our new double-cold experimental design is more appropriate for emergent DTI challenges, (3) that our novel MUSDTI meta-model outperforms SOTA models, (4) that RP can improve upon individual models as an ensembling method, and finally, (5) RP can be utilized for low computation transfer learning. This work introduces a number of important revelations for the field of DTI prediction and sequence-based, pairwise prediction in general.

A cross-sectional study of COVID-19 vaccination patterns among patients with epilepsy in Hong Kong.

OBJECTIVE: As Hong Kong faced the 5th wave of the COVID-19 pandemic, the facilitators and hurdles toward effective vaccination is important for healthcare professionals to understand the vaccination gap among patients with epilepsy. METHODS: A cross-sectional, pragmatic study of COVID-19 vaccination was performed at a tertiary epilepsy center with regards to patterns of vaccination and any unusually high rate of adverse events. Patients having recent visits at the epilepsy center (4 months) had their anonymized electronic linkage records examined 12 months after the inception of vaccination program for types of vaccines, seizure demographics, and adverse events following immunization (AEFI). RESULTS: A total of 200 patients with epilepsy and their anonymized data were analyzed. The vaccine uptake was approximately 60% of that of the general population. Twice as many patients with epilepsy chose to receive mRNA vaccine as compared with inactivated vaccine. The proportion of patients who kept up-to-date with all available dosing was 7%. Patients with epilepsy with genetic etiology were least likely to receive vaccination (13/38, 34%, P = .02). There was no unreasonably high rate of unacceptable side effects after vaccination among patients with epilepsy. Only 3 patients reported worsening of seizures without meeting the criteria for AEFI. Refractory epilepsy, allergy to antiseizure medications and elder age (≥65) did not confer any significant difference in vaccination patterns or adverse effects. SIGNIFICANCE: A vaccination gap exists among epilepsy patients which calls for actionable strategies for improving vaccine uptake, including education and outreach programs.

Pandemic of the aging society - sporadic cerebral small vessel disease.

ABSTRACT: Age-related sporadic cerebral small vessel disease (CSVD) has gained increasing attention over the past decades because of its increasing prevalence associated with an aging population. The widespread application of and advances in brain magnetic resonance imaging in recent decades have significantly increased researchers' understanding in the in vivo evolution of CSVD, its impact upon the brain, its risk factors, and the mechanisms that explain the various clinical manifestation associated with sporadic CSVD. In this review, we aimed to provide an update on the pathophysiology, risk factors, biomarkers, and the determinants and spectrum of the clinical manifestation of sporadic CSVD.

International Survey of Mechanical Thrombectomy Stroke Systems of Care During COVID-19 Pandemic.

BACKGROUND: The COVID-19 pandemic has strained the healthcare systems across the world but its impact on acute stroke care is just being elucidated. We hypothesized a major global impact of COVID-19 not only on stroke volumes but also on various aspects of thrombectomy systems. AIMS: We conducted a convenience electronic survey with a 21-item questionnaire aimed to identify the changes in stroke admission volumes and thrombectomy treatment practices seen during a specified time period of the COVID-19 pandemic. METHODS: The survey was designed using Qualtrics software and sent to stroke and neuro-interventional physicians around the world who are part of the Global Executive Committee (GEC) of Mission Thrombectomy 2020, a global coalition under the aegis of Society of Vascular and Interventional Neurology, between April 5th and May 15th, 2020. RESULTS: There were 113 responses to the survey across 25 countries with a response rate of 31% among the GEC members. Globally there was a median 33% decrease in stroke admissions and a 25% decrease in mechanical thrombectomy (MT) procedures during the COVID-19 pandemic period until May 15th, 2020 compared to pre-pandemic months. The intubation policy for MT procedures during the pandemic was highly variable across participating centers: 44% preferred intubating all patients, including 25% of centers that changed their policy to preferred-intubation (PI) from preferred non-intubation (PNI). On the other hand, 56% centers preferred not intubating patients undergoing MT, which included 27% centers that changed their policy from PI to PNI. There was no significant difference in rate of COVID-19 infection between PI versus PNI centers (p=0.60) or if intubation policy was changed in either direction (p=1.00). Low-volume (<10 stroke/month) compared with high-volume stroke centers (>20 strokes/month) were less likely to have neurointerventional suite specific written personal protective equipment protocols (74% vs 88%) and if present, these centers were more likely to report them to be inadequate (58% vs 92%). CONCLUSION: Our data provides a comprehensive snapshot of the impact on acute stroke care observed worldwide during the pandemic. Overall, respondents reported decreased stroke admissions as well as decreased cases of MT with no clear preponderance in intubation policy during MT. DATA ACCESS STATEMENT: The corresponding author will consider requests for sharing survey data. The study was exempt from institutional review board approval as it did not involve patient level data.