Development, External Validation, and Biomolecular Corroboration of Interoperable Models for Identifying Critically Ill Children at Risk of Neurologic Morbidity.

Importance: Declining mortality in the field of pediatric critical care medicine has shifted practicing clinicians' attention to preserving patients' neurodevelopmental potential as a main objective. Earlier identification of critically ill children at risk for incurring neurologic morbidity would facilitate heightened surveillance that could lead to timelier clinical detection, earlier interventions, and preserved neurodevelopmental trajectory. Objective: Develop machine-learning models for identifying acquired neurologic morbidity while hospitalized with critical illness and assess correlation with contemporary serum-based, brain injury-derived biomarkers. Design: Retrospective cohort study. Setting: Two large, quaternary children's hospitals. Exposures: Critical illness. Main Outcomes and Measures: The outcome was neurologic morbidity, defined according to a computable, composite definition at the development site or an order for neurocritical care consultation at the validation site. Models were developed using varying time windows for temporal feature engineering and varying censored time horizons prior to identified neurologic morbidity. Optimal models were selected based on F1 scores, cohort sizes, calibration, and data availability for eventual deployment. A generalizable created at the development site was assessed at an external validation site and optimized with spline recalibration. Correlation was assessed between development site model predictions and measurements of brain biomarkers from a convenience cohort. Results: After exclusions there were 14,222-25,171 encounters from 2010-2022 in the development site cohorts and 6,280-6,373 from 2018-2021 in the validation site cohort. At the development site, an extreme gradient boosted model (XGBoost) with a 12-hour time horizon and 48-hour feature engineering window had an F1-score of 0.54, area under the receiver operating characteristics curve (AUROC) of 0.82, and a number needed to alert (NNA) of 2. A generalizable XGBoost model with a 24-hour time horizon and 48-hour feature engineering window demonstrated an F1-score of 0.37, AUROC of 0.81, AUPRC of 0.51, and NNA of 4 at the validation site. After recalibration at the validation site, the Brier score was 0.04. Serum levels of the brain injury biomarker glial fibrillary acidic protein measurements significantly correlated with model output (rs=0.34; P=0.007). Conclusions and Relevance: We demonstrate a well-performing ensemble of models for predicting neurologic morbidity in children with biomolecular corroboration. Prospective assessment and refinement of biomarker-coupled risk models in pediatric critical illness is warranted.

Broadening the capture of natural products mentioned in FAERS using fuzzy string-matching and a Siamese neural network.

Increased sales of natural products (NPs) in the US and growing safety concerns highlight the need for NP pharmacovigilance. A challenge for NP pharmacovigilance is ambiguity when referring to NPs in spontaneous reporting systems. We used a combination of fuzzy string-matching and a neural network to reduce this ambiguity. Our aim is to increase the capture of reports involving NPs in the US Food and Drug Administration Adverse Event Reporting System (FAERS). For this, we utilized Gestalt pattern-matching (GPM) and Siamese neural network (SM) to identify potential mentions of NPs of interest in 389,386 FAERS reports with unmapped drug names. A team of health professionals refined the candidates identified in the previous step through manual review and annotation. After candidate adjudication, GPM identified 595 unique NP names and SM 504. There was little overlap between candidates identified by each (Non-overlapping: GPM 347, SM 248). We identified a total of 686 novel NP names from FAERS reports. Including these names in the FAERS collection yielded 3,486 additional reports mentioning NPs.

Broadening the Capture of Natural Products Mentioned in FAERS Using Fuzzy String-Matching and a Siamese Neural Network.

Increased sales of natural products (NPs) in the US and growing safety concerns highlight the need for NP pharmacovigilance. A challenge for NP pharmacovigilance is ambiguity when referring to NPs in spontaneous reporting systems. We used a combination of fuzzy string-matching and a neural network to reduce this ambiguity. We aim to increase the capture of reports involving NPs in the US Food and Drug Administration Adverse Event Reporting System (FAERS). Gestalt pattern-matching (GPM) and Siamese neural network (SM) were used to identify potential mentions of NPs of interest in 389,386 FAERS reports with unmapped drug names. We refined the identified candidates through manual review and annotation by health professionals. After adjudication, GPM identified 595 unique NP names and SM 504. There was little overlap between candidates identified by the approaches (Non-overlapping: GPM 347, SM 248). In total, 686 novel NP names were identified in the unmapped FAERS reports. Including these names in the FAERS collection yielded 3,486 additional reports mentioning NPs.

Unsupervised Mining of HLA-I Peptidomes Reveals New Binding Motifs and Potential False Positives in the Community Database.

Modern vaccine designs and studies of human leukocyte antigen (HLA)-mediated immune responses rely heavily on the knowledge of HLA allele-specific binding motifs and computational prediction of HLA-peptide binding affinity. Breakthroughs in HLA peptidomics have considerably expanded the databases of natural HLA ligands and enabled detailed characterizations of HLA-peptide binding specificity. However, cautions must be made when analyzing HLA peptidomics data because identified peptides may be contaminants in mass spectrometry or may weakly bind to the HLA molecules. Here, a hybrid de novo peptide sequencing approach was applied to large-scale mono-allelic HLA peptidomics datasets to uncover new ligands and refine current knowledge of HLA binding motifs. Up to 12-40% of the peptidomics data were low-binding affinity peptides with an arginine or a lysine at the C-terminus and likely to be tryptic peptide contaminants. Thousands of these peptides have been reported in a community database as legitimate ligands and might be erroneously used for training prediction models. Furthermore, unsupervised clustering of identified ligands revealed additional binding motifs for several HLA class I alleles and effectively isolated outliers that were experimentally confirmed to be false positives. Overall, our findings expanded the knowledge of HLA binding specificity and advocated for more rigorous interpretation of HLA peptidomics data that will ensure the high validity of community HLA ligandome databases.