Projections of Single-level and Multilevel Spinal Instrumentation Procedure Volume and Associated Costs for Medicare Patients to 2050.

BACKGROUND: Instrumented spinal fusions can be used in the treatment of vertebral fractures, spinal instability, and scoliosis or kyphosis. Construct-level selection has notable implications on postoperative recovery, alignment, and mobility. This study sought to project future trends in the implementation rates and associated costs of single-level versus multilevel instrumentation procedures in US Medicare patients aged older than 65 years in the United States. METHODS: Data were acquired from the Centers for Medicare & Medicaid Services from January 1, 2000, to December 31, 2019. Procedure costs and counts were abstracted using Current Procedural Terminology codes to identify spinal level involvement. The Prophet machine learning algorithm was used, using a Bayesian Inference framework, to generate point forecasts for 2020 to 2050 and 95% forecast intervals (FIs). Sensitivity analyses were done by comparing projections from linear, log-linear, Poisson and negative-binomial, and autoregressive integrated moving average models. Costs were adjusted for inflation using the 2019 US Bureau of Labor Statistics' Consumer Price Index. RESULTS: Between 2000 and 2019, the annual spinal instrumentation volume increased by 776% (from 7,342 to 64,350 cases) for single level, by 329% (from 20,319 to 87,253 cases) for two-four levels, by 1049% (from 1,218 to 14,000 cases) for five-seven levels, and by 739% (from 193 to 1,620 cases) for eight-twelve levels (P < 0.0001). The inflation-adjusted reimbursement for single-level instrumentation procedures decreased 45.6% from $1,148.15 to $788.62 between 2000 and 2019, which is markedly lower than for other prevalent orthopaedic procedures: total shoulder arthroplasty (-23.1%), total hip arthroplasty (-39.2%), and total knee arthroplasty (-42.4%). By 2050, the number of single-level spinal instrumentation procedures performed yearly is projected to be 124,061 (95% FI, 87,027 to 142,907), with associated costs of $93,900,672 (95% FI, $80,281,788 to $108,220,932). CONCLUSIONS: The number of single-level instrumentation procedures is projected to double by 2050, while the number of two-four level procedures will double by 2040. These projections offer a measurable basis for resource allocation and procedural distribution.

Dry-Cleaning Chemicals and a Cluster of Parkinson's Disease and Cancer: A Retrospective Investigation.

BACKGROUND: Environmental exposure to trichloroethylene (TCE), a carcinogenic dry-cleaning chemical, may be linked to Parkinson's disease (PD). OBJECTIVE: The objective of this study was to determine whether PD and cancer were elevated among attorneys who worked near a contaminated site. METHODS: We surveyed and evaluated attorneys with possible exposure and assessed a comparison group. RESULTS: Seventy-nine of 82 attorneys (96.3%; mean [SD] age: 69.5 [11.4] years; 89.9% men) completed at least one phase of the study. For comparison, 75 lawyers (64.9 [10.2] years; 65.3% men) underwent clinical evaluations. Four (5.1%) of them who worked near the polluted site reported PD, more than expected based on age and sex (1.7%; P = 0.01) but not significantly higher than the comparison group (n = 1 [1.3%]; P = 0.37). Fifteen (19.0%), compared to four in the comparison group (5.3%; P = 0.049), had a TCE-related cancer. CONCLUSIONS: In a retrospective study, diagnoses of PD and TCE-related cancers appeared to be elevated among attorneys who worked next to a contaminated dry-cleaning site. © 2024 The Authors. Movement Disorders published by Wiley Periodicals LLC on behalf of International Parkinson and Movement Disorder Society.

Generating Molecular Fragmentation Graphs with Autoregressive Neural Networks.

The accurate prediction of tandem mass spectra from molecular structures has the potential to unlock new metabolomic discoveries by augmenting the community's libraries of experimental reference standards. Cheminformatic spectrum prediction strategies use a "bond-breaking" framework to iteratively simulate mass spectrum fragmentations, but these methods are (a) slow due to the need to exhaustively and combinatorially break molecules and (b) inaccurate as they often rely upon heuristics to predict the intensity of each resulting fragment; neural network alternatives mitigate computational cost but are black-box and not inherently more accurate. We introduce a physically grounded neural approach that learns to predict each breakage event and score the most relevant subset of molecular fragments quickly and accurately. We evaluate our model by predicting spectra from both public and private standard libraries, demonstrating that our hybrid approach offers state-of-the-art prediction accuracy, improved metabolite identification from a database of candidates, and higher interpretability when compared to previous breakage methods and black-box neural networks. The grounding of our approach in physical fragmentation events shows especially great promise for elucidating natural product molecules with more complex scaffolds.

Anterior cervical osteotomy of diffuse idiopathic skeletal hyperostosis lesions with computer-assisted navigation surgery: A case report.

Key Clinical Message: Diffuse idiopathic skeletal hyperostosis (DISH) involves spine ligament ossification. Computer-assisted navigation (CAN) effectively aids complex surgeries, such as anterior cervical osteotomy, to alleviate progressive DISH-related dysphagia. Abstract: We describe a 68-year-old man with sudden onset dysphagia to both solids and liquids. Radiographic Imaging revealed DISH lesions from C2 down to the thoracic spine. The patient was successfully treated with CAN anterior osteotomy and resection of DISH lesions from C3-C6 and had complete symptom relief within 2 weeks post-operatively.

MIST-CF: Chemical Formula Inference from Tandem Mass Spectra.

Chemical formula annotation for tandem mass spectrometry (MS/MS) data is the first step toward structurally elucidating unknown metabolites. While great strides have been made toward solving this problem, the current state-of-the-art method depends on time-intensive, proprietary, and expert-parametrized fragmentation tree construction and scoring. In this work, we extend our previous spectrum Transformer methodology into an energy-based modeling framework, MIST-CF: Metabolite Inference with Spectrum Transformers for Chemical Formula prediction, for learning to rank chemical formula and adduct assignments given an unannotated MS/MS spectrum. Importantly, MIST-CF learns in a data-dependent fashion using a Formula Transformer neural network architecture and circumvents the need for fragmentation tree construction. We train and evaluate our model on a large open-access database, showing an absolute improvement of 10% top 1 accuracy over other neural network architectures. We further validate our approach on the CASMI2022 challenge data set, achieving nearly equivalent performance to the winning entry within the positive mode category without any manual curation or postprocessing of our results. These results demonstrate an exciting strategy to more powerfully leverage MS2 fragment peaks for predicting MS1 precursor chemical formulas with data-driven learning.

Prospective Comparison of Two Robotically Navigated Pedicle Screw Instrumentation Techniques.

This study aimed to compare screw accuracy and incidence of skive between two robotically navigated instrumented techniques in posterior spine fusion surgery: manual anti-skive instrumentation with an anti-skive cannula (ASC) and the use of a navigated, high-speed drill (HSD). Over a 3-year period, consecutive patients are undergoing RNA posterior fusion surgery with either ASC (n = 53) or HSD (n = 63). Both groups met a value of approximately 292 screws in our analysis (296 ASC, 294 HSD), which was determined by a biostatistician at an academic institution. Screw accuracy and skive was analyzed using preoperative CT and intraoperative three-dimensional (3D) fluoroscopy. Among 590 planned robotically inserted pedicle screws (296 ASC, 294 HSD), 245 ASC screws (82.8%) and 283 HSD screws (96.3%) were successfully inserted (p < 0.05). Skive events occurred in 4/283 (1.4%) HSD screws and 15/245 (6.2%) ASC screws (p < 0.05). HSD screws showed better accuracy in the axial and sagittal planes, being closer to planned trajectories in all directions except cranial deviation (p < 0.05). Additionally, HSD had a significantly lower time per screw (1.9 ± 1.0 min) compared to ASC (3.2 ± 2.0 min, p < 0.001). No adverse clinical effects were observed. The HSD technique showed significant improvements in time and screw accuracy compared to ASC. Biplanar fluoroscopy and 3D imaging resulted in significantly lower radiation exposure and time compared to ASC. These significant findings in the HSD group may be attributed to the lower occurrence of malpositioned screws, leading to a decrease in the need for second authentication. This represents a notable iterative improvement of the RNA platform.

Externally validated deep learning model to identify prodromal Parkinson's disease from electrocardiogram.

Little is known about electrocardiogram (ECG) markers of Parkinson's disease (PD) during the prodromal stage. The aim of the study was to build a generalizable ECG-based fully automatic artificial intelligence (AI) model to predict PD risk during the prodromal stage, up to 5 years before disease diagnosis. This case-control study included samples from Loyola University Chicago (LUC) and University of Tennessee-Methodist Le Bonheur Healthcare (MLH). Cases and controls were matched according to specific characteristics (date, age, sex and race). Clinical data were available from May, 2014 onward at LUC and from January, 2015 onward at MLH, while the ECG data were available as early as 1990 in both institutes. PD was denoted by at least two primary diagnostic codes (ICD9 332.0; ICD10 G20) at least 30 days apart. PD incidence date was defined as the earliest of first PD diagnostic code or PD-related medication prescription. ECGs obtained at least 6 months before PD incidence date were modeled to predict a subsequent diagnosis of PD within three time windows: 6 months-1 year, 6 months-3 years, and 6 months-5 years. We applied a novel deep neural network using standard 10-s 12-lead ECGs to predict PD risk at the prodromal phase. This model was compared to multiple feature engineering-based models. Subgroup analyses for sex, race and age were also performed. Our primary prediction model was a one-dimensional convolutional neural network (1D-CNN) that was built using 131 cases and 1058 controls from MLH, and externally validated on 29 cases and 165 controls from LUC. The model was trained on 90% of the MLH data, internally validated on the remaining 10% and externally validated on LUC data. The best performing model resulted in an external validation AUC of 0.67 when predicting future PD at any time between 6 months and 5 years after the ECG. Accuracy increased when restricted to ECGs obtained within 6 months to 3 years before PD diagnosis (AUC 0.69) and was highest when predicting future PD within 6 months to 1 year (AUC 0.74). The 1D-CNN model based on raw ECG data outperformed multiple models built using more standard ECG feature engineering approaches. These results demonstrate that a predictive model developed in one cohort using only raw 10-s ECGs can effectively classify individuals with prodromal PD in an independent cohort, particularly closer to disease diagnosis. Standard ECGs may help identify individuals with prodromal PD for cost-effective population-level early detection and inclusion in disease-modifying therapeutic trials.

Background, techniques, applications, current trends, and future directions of minimally invasive endoscopic spine surgery: A review of literature.

Across many of the surgical specialties, the use of minimally invasive techniques that utilize indirect visualization has been increasingly replacing traditional techniques which utilize direct visualization. Arthroscopic surgery of the appendicular skeleton has evolved dramatically and become an integral part of musculoskeletal surgery over the last several decades, allowing surgeons to achieve similar or better outcomes, while reducing cost and recovery time. However, to date, the axial skeleton, with its close proximity to critical neural and vascular structures, has not adopted endoscopic techniques at as rapid of a rate. Over the past decade, increased patient demand for less invasive spine surgery combined with surgeon desire to meet these demands has driven significant evolution and innovation in endoscopic spine surgery. In addition, there has been an enormous advancement in technologies that assist in navigation and automation that help surgeons circumvent limitations of direct visualization inherent to less invasive techniques. There are currently a multitude of endoscopic techniques and approaches that can be utilized in the treatment of spine disorders, many of which are evolving rapidly. Here we present a review of the field of endoscopic spine surgery, including the background, techniques, applications, current trends, and future directions, to help providers gain a better understanding of this growing modality in spine surgery.

Risk of Parkinson Disease Among Service Members at Marine Corps Base Camp Lejeune.

Importance: An increased risk of Parkinson disease (PD) has been associated with exposure to the solvent trichloroethylene (TCE), but data are limited. Millions of people in the US and worldwide are exposed to TCE in air, food, and water. Objective: To test whether the risk of PD is higher in veterans who served at Marine Corps Base Camp Lejeune, whose water supply was contaminated with TCE and other volatile organic compounds (VOCs), compared with veterans who did not serve on that base. Design, Setting, and Participants: This population-based cohort study examined the risk for PD among all Marines and Navy personnel who resided at Camp Lejeune, North Carolina (contaminated water) (n = 172 128), or Camp Pendleton, California (uncontaminated water) (n = 168 361), for at least 3 months between 1975 and 1985, with follow-up from January 1, 1997, until February 17, 2021. Veterans Health Administration and Medicare databases were searched for International Classification of Diseases diagnostic codes for PD or other forms of parkinsonism and related medications and for diagnostic codes indicative of prodromal disease. Parkinson disease diagnoses were confirmed by medical record review. Exposures: Water supplies at Camp Lejeune were contaminated with several VOCs. Levels were highest for TCE, with monthly median values greater than 70-fold the permissible amount. Main Outcome and Measures: Risk of PD in former residents of Camp Lejeune relative to residents of Camp Pendleton. In those without PD or another form of parkinsonism, the risk of being diagnosed with features of prodromal PD were assessed individually and cumulatively using likelihood ratio tests. Results: Health data were available for 158 122 veterans (46.4%). Demographic characteristics were similar between Camp Lejeune (5.3% women, 94.7% men; mean [SD] attained age of 59.64 [4.43] years; 29.7% Black, 6.0% Hispanic, 67.6% White; and 2.7% other race and ethnicity) and Camp Pendleton (3.8% women, 96.2% men; mean [SD] age, 59.80 [4.62] years; 23.4% Black, 9.4% Hispanic, 71.1% White, and 5.5% other race and ethnicity). A total of 430 veterans had PD, with 279 from Camp Lejeune (prevalence, 0.33%) and 151 from Camp Pendleton (prevalence, 0.21%). In multivariable models, Camp Lejeune veterans had a 70% higher risk of PD (odds ratio, 1.70; 95% CI, 1.39-2.07; P < .001). No excess risk was found for other forms of neurodegenerative parkinsonism. Camp Lejeune veterans also had a significantly increased risk of prodromal PD diagnoses, including tremor, anxiety, and erectile dysfunction, and higher cumulative prodromal risk scores. Conclusions and Relevance: The study's findings suggest that the risk of PD is higher in persons exposed to TCE and other VOCs in water 4 decades ago. Millions worldwide have been and continue to be exposed to this ubiquitous environmental contaminant.

Early detection of Parkinson's disease through enriching the electronic health record using a biomedical knowledge graph.

Introduction: Early diagnosis of Parkinson's disease (PD) is important to identify treatments to slow neurodegeneration. People who develop PD often have symptoms before the disease manifests and may be coded as diagnoses in the electronic health record (EHR). Methods: To predict PD diagnosis, we embedded EHR data of patients onto a biomedical knowledge graph called Scalable Precision medicine Open Knowledge Engine (SPOKE) and created patient embedding vectors. We trained and validated a classifier using these vectors from 3,004 PD patients, restricting records to 1, 3, and 5 years before diagnosis, and 457,197 non-PD group. Results: The classifier predicted PD diagnosis with moderate accuracy (AUC = 0.77 ± 0.06, 0.74 ± 0.05, 0.72 ± 0.05 at 1, 3, and 5 years) and performed better than other benchmark methods. Nodes in the SPOKE graph, among cases, revealed novel associations, while SPOKE patient vectors revealed the basis for individual risk classification. Discussion: The proposed method was able to explain the clinical predictions using the knowledge graph, thereby making the predictions clinically interpretable. Through enriching EHR data with biomedical associations, SPOKE may be a cost-efficient and personalized way to predict PD diagnosis years before its occurrence.

Robotic-navigated assistance in spine surgery.

The aim of this study was to assess the accuracy of pedicle screw placement, as well as intraoperative factors, radiation exposure, and complication rates in adult patients with degenerative disorders of the thoracic and lumbar spines who have undergone robotic-navigated spinal surgery using a contemporary system. The authors reviewed the prospectively collected data on 196 adult patients who had pedicle screws implanted with robot-navigated assistance (RNA) using the Mazor X Stealth system between June 2019 and March 2022. Pedicle screws were implanted by one experienced spinal surgeon after completion of a learning period. The accuracy of pedicle screw placement was determined using intraoperative 3D fluoroscopy. A total of 1,123 pedicle screws were implanted: 1,001 screws (89%) were placed robotically, 63 (6%) were converted from robotic placement to a freehand technique, and 59 (5%) were planned to be implanted freehand. Of the robotically placed screws, 942 screws (94%) were determined to be Gertzbein and Robbins grade A with median deviation of 0.8 mm (interquartile range 0.4 to 1.6). Skive events were noted with 20 pedicle screws (1.8%). No adverse clinical sequelae were noted in the 90-day follow-up. The mean fluoroscopic exposure per screw was 4.9 seconds (SD 3.8). RNA is highly accurate and reliable, with a low rate of abandonment once mastered. No adverse clinical sequelae occurred after implanting a large series of pedicle screws using the latest generation of RNA. Understanding of patient-specific anatomical features and the real-time intraoperative identification of risk factors for suboptimal screw placement have the potential to improve accuracy further.

Trichloroethylene: An Invisible Cause of Parkinson's Disease?

The etiologies of Parkinson's disease (PD) remain unclear. Some, such as certain genetic mutations and head trauma, are widely known or easily identified. However, these causes or risk factors do not account for the majority of cases. Other, less visible factors must be at play. Among these is a widely used industrial solvent and common environmental contaminant little recognized for its likely role in PD: trichloroethylene (TCE). TCE is a simple, six-atom molecule that can decaffeinate coffee, degrease metal parts, and dry clean clothes. The colorless chemical was first linked to parkinsonism in 1969. Since then, four case studies involving eight individuals have linked occupational exposure to TCE to PD. In addition, a small epidemiological study found that occupational or hobby exposure to the solvent was associated with a 500% increased risk of developing PD. In multiple animal studies, the chemical reproduces the pathological features of PD.Exposure is not confined to those who work with the chemical. TCE pollutes outdoor air, taints groundwater, and contaminates indoor air. The molecule, like radon, evaporates from underlying soil and groundwater and enters homes, workplaces, or schools, often undetected. Despite widespread contamination and increasing industrial, commercial, and military use, clinical investigations of TCE and PD have been limited. Here, through a literature review and seven illustrative cases, we postulate that this ubiquitous chemical is contributing to the global rise of PD and that TCE is one of its invisible and highly preventable causes. Further research is now necessary to examine this hypothesis.

Resonant learning in scale-free networks.

Large networks of interconnected components, such as genes or machines, can coordinate complex behavioral dynamics. One outstanding question has been to identify the design principles that allow such networks to learn new behaviors. Here, we use Boolean networks as prototypes to demonstrate how periodic activation of network hubs provides a network-level advantage in evolutionary learning. Surprisingly, we find that a network can simultaneously learn distinct target functions upon distinct hub oscillations. We term this emergent property resonant learning, as the new selected dynamical behaviors depend on the choice of the period of the hub oscillations. Furthermore, this procedure accelerates the learning of new behaviors by an order of magnitude faster than without oscillations. While it is well-established that modular network architecture can be selected through evolutionary learning to produce different network behaviors, forced hub oscillations emerge as an alternative evolutionary learning strategy for which network modularity is not necessarily required.

Appendicitis Following Blunt Abdominal Trauma: An Illustrative Case.

Acute appendicitis is the most common abdominal surgical emergency, with an average of 7-9% of individuals developing the condition within their lifetime.1 While cases of acute traumatic appendicitis are rare, medical literature supports their plausibility with the most famous case stretching back to the controversial 1926 death of stunt performer, Harry Houdini. Several mechanisms have been proposed by which blunt abdominal trauma results in acute appendicitis. In this review, we describe a young, otherwise healthy male, who developed epigastric abdominal pain after being struck in the abdomen while wrestling with his cousin of similar age. The patient was found to have peri-appendiceal inflammatory change, appendiceal mural thickening and edema consistent with acute uncomplicated appendicitis.

Machine learning modeling of family wide enzyme-substrate specificity screens.

Biocatalysis is a promising approach to sustainably synthesize pharmaceuticals, complex natural products, and commodity chemicals at scale. However, the adoption of biocatalysis is limited by our ability to select enzymes that will catalyze their natural chemical transformation on non-natural substrates. While machine learning and in silico directed evolution are well-posed for this predictive modeling challenge, efforts to date have primarily aimed to increase activity against a single known substrate, rather than to identify enzymes capable of acting on new substrates of interest. To address this need, we curate 6 different high-quality enzyme family screens from the literature that each measure multiple enzymes against multiple substrates. We compare machine learning-based compound-protein interaction (CPI) modeling approaches from the literature used for predicting drug-target interactions. Surprisingly, comparing these interaction-based models against collections of independent (single task) enzyme-only or substrate-only models reveals that current CPI approaches are incapable of learning interactions between compounds and proteins in the current family level data regime. We further validate this observation by demonstrating that our no-interaction baseline can outperform CPI-based models from the literature used to guide the discovery of kinase inhibitors. Given the high performance of non-interaction based models, we introduce a new structure-based strategy for pooling residue representations across a protein sequence. Altogether, this work motivates a principled path forward in order to build and evaluate meaningful predictive models for biocatalysis and other drug discovery applications.

Remote smartphone monitoring of Parkinson's disease and individual response to therapy.

Remote health assessments that gather real-world data (RWD) outside clinic settings require a clear understanding of appropriate methods for data collection, quality assessment, analysis and interpretation. Here we examine the performance and limitations of smartphones in collecting RWD in the remote mPower observational study of Parkinson's disease (PD). Within the first 6 months of study commencement, 960 participants had enrolled and performed at least five self-administered active PD symptom assessments (speeded tapping, gait/balance, phonation or memory). Task performance, especially speeded tapping, was predictive of self-reported PD status (area under the receiver operating characteristic curve (AUC) = 0.8) and correlated with in-clinic evaluation of disease severity (r = 0.71; P < 1.8 × 10-6) when compared with motor Movement Disorder Society-Unified Parkinson's Disease Rating Scale (MDS-UPDRS). Although remote assessment requires careful consideration for accurate interpretation of RWD, our results support the use of smartphones and wearables in objective and personalized disease assessments.

Preventing Parkinson's Disease: An Environmental Agenda.

Fueled by aging populations and continued environmental contamination, the global burden of Parkinson's disease (PD) is increasing. The disease, or more appropriately diseases, have multiple environmental and genetic influences but no approved disease modifying therapy. Additionally, efforts to prevent this debilitating disease have been limited. As numerous environmental contaminants (e.g., pesticides, metals, industrial chemicals) are implicated in PD, disease prevention is possible. To reduce the burden of PD, we have compiled preclinical and clinical research priorities that highlight both disease prediction and primary prevention. Though not exhaustive, the "PD prevention agenda" builds upon many years of research by our colleagues and proposes next steps through the lens of modifiable risk factors. The agenda identifies ten specific areas of further inquiry and considers the funding and policy changes that will be necessary to help prevent the world's fastest growing brain disease.

Predicting Parkinson's Disease and Its Pathology via Simple Clinical Variables.

BACKGROUND: Parkinson's disease (PD) is a chronic, disabling neurodegenerative disorder. OBJECTIVE: To predict a future diagnosis of PD using questionnaires and simple non-invasive clinical tests. METHODS: Participants in the prospective Kuakini Honolulu-Asia Aging Study (HAAS) were evaluated biannually between 1995-2017 by PD experts using standard diagnostic criteria. Autopsies were sought on all deaths. We input simple clinical and risk factor variables into an ensemble-tree based machine learning algorithm and derived models to predict the probability of developing PD. We also investigated relationships of predictive models and neuropathologic features such as nigral neuron density. RESULTS: The study sample included 292 subjects, 25 of whom developed PD within 3 years and 41 by 5 years. 116 (46%) of 251 subjects not diagnosed with PD underwent autopsy. Light Gradient Boosting Machine modeling of 12 predictors correctly classified a high proportion of individuals who developed PD within 3 years (area under the curve (AUC) 0.82, 95%CI 0.76-0.89) or 5 years (AUC 0.77, 95%CI 0.71-0.84). A large proportion of controls who were misclassified as PD had Lewy pathology at autopsy, including 79%of those who died within 3 years. PD probability estimates correlated inversely with nigral neuron density and were strongest in autopsies conducted within 3 years of index date (r = -0.57, p < 0.01). CONCLUSION: Machine learning can identify persons likely to develop PD during the prodromal period using questionnaires and simple non-invasive tests. Correlation with neuropathology suggests that true model accuracy may be considerably higher than estimates based solely on clinical diagnosis.

EHreact: Extended Hasse Diagrams for the Extraction and Scoring of Enzymatic Reaction Templates.

Data-driven computer-aided synthesis planning utilizing organic or biocatalyzed reactions from large databases has gained increasing interest in the last decade, sparking the development of numerous tools to extract, apply, and score general reaction templates. The generation of reaction rules for enzymatic reactions is especially challenging since substrate promiscuity varies between enzymes, causing the optimal levels of rule specificity and optimal number of included atoms to differ between enzymes. This complicates an automated extraction from databases and has promoted the creation of manually curated reaction rule sets. Here, we present EHreact, a purely data-driven open-source software tool, to extract and score reaction rules from sets of reactions known to be catalyzed by an enzyme at appropriate levels of specificity without expert knowledge. EHreact extracts and groups reaction rules into tree-like structures, Hasse diagrams, based on common substructures in the imaginary transition structures. Each diagram can be utilized to output a single or a set of reaction rules, as well as calculate the probability of a new substrate to be processed by the given enzyme by inferring information about the reactive site of the enzyme from the known reactions and their grouping in the template tree. EHreact heuristically predicts the activity of a given enzyme on a new substrate, outperforming current approaches in accuracy and functionality.

Evidential Deep Learning for Guided Molecular Property Prediction and Discovery.

While neural networks achieve state-of-the-art performance for many molecular modeling and structure-property prediction tasks, these models can struggle with generalization to out-of-domain examples, exhibit poor sample efficiency, and produce uncalibrated predictions. In this paper, we leverage advances in evidential deep learning to demonstrate a new approach to uncertainty quantification for neural network-based molecular structure-property prediction at no additional computational cost. We develop both evidential 2D message passing neural networks and evidential 3D atomistic neural networks and apply these networks across a range of different tasks. We demonstrate that evidential uncertainties enable (1) calibrated predictions where uncertainty correlates with error, (2) sample-efficient training through uncertainty-guided active learning, and (3) improved experimental validation rates in a retrospective virtual screening campaign. Our results suggest that evidential deep learning can provide an efficient means of uncertainty quantification useful for molecular property prediction, discovery, and design tasks in the chemical and physical sciences.