Development of Electronic Health Record-Based Machine Learning Models to Predict Barrett's Esophagus and Esophageal Adenocarcinoma Risk.

INTRODUCTION: Screening for Barrett's esophagus (BE) is suggested in those with risk factors, but remains underutilized. BE/esophageal adenocarcinoma (EAC) risk prediction tools integrating multiple risk factors have been described. However, accuracy remains modest (area under the receiver-operating curve [AUROC] ≤0.7), and clinical implementation has been challenging. We aimed to develop machine learning (ML) BE/EAC risk prediction models from an electronic health record (EHR) database. METHODS: The Clinical Data Analytics Platform, a deidentified EHR database of 6 million Mayo Clinic patients, was used to predict BE and EAC risk. BE and EAC cases and controls were identified using International Classification of Diseases codes and augmented curation (natural language processing) techniques applied to clinical, endoscopy, laboratory, and pathology notes. Cases were propensity score matched to 5 independent randomly selected control groups. An ensemble transformer-based ML model architecture was used to develop predictive models. RESULTS: We identified 8,476 BE cases, 1,539 EAC cases, and 252,276 controls. The BE ML transformer model had an overall sensitivity, specificity, and AUROC of 76%, 76%, and 0.84, respectively. The EAC ML transformer model had an overall sensitivity, specificity, and AUROC of 84%, 70%, and 0.84, respectively. Predictors of BE and EAC included conventional risk factors and additional novel factors, such as coronary artery disease, serum triglycerides, and electrolytes. DISCUSSION: ML models developed on an EHR database can predict incident BE and EAC risk with improved accuracy compared with conventional risk factor-based risk scores. Such a model may enable effective implementation of a minimally invasive screening technology.

In vitro plant regeneration and genetic transformation of Dichanthium annulatum.

Optimization of in vitro plant regeneration and genetic transformation of apomictic species such as Dichanthium annulatum would enable transfer of desirable genes. Seven genotypes of this grass species were screened through mature seed explant for embryogenic callus induction, callus growth and quality (color and texture), and shoot induction. Genotype IG-1999, which produced highly embryogenic, rapidly growing good-quality callus capable of regenerating at a high frequency, was selected for transformation experiments. Using a binary vector (pCAMBIA1305), frequency of GUS expression was compared between two methods of transformation. Bombardment of embryogenic calli with gold particles coated with pCAMBIA1305 at a distance of 11 cm, pressure of 4 bars, and vacuum of 27 Hg passing through 100 muM mesh produced maximum GUS expression (23%). Agrobacterium infection was maximum at an optical density of 2.0 when cocultured under vacuum for 15 min and cocultivated for 3 days at 28 degrees C in constant dark on MS medium of pH 5.8 with 3 mg/l 2,4-D, and 400 muM acetosyringone. Among two binary vectors used for Agrobacterium-mediated transformation, pCAMBIA1301 showed higher frequency of GUS expression while pCAMBIA1305 recorded more of the GUS spots per callus. Supplementation of acetosyringone in the cocultivation medium was found indispensable for Agrobacterium-mediated transformation. Injuring the calli through gold particle bombardment before their cocultivation with Agrobacterium improved the transformation efficiency. Several transgenic plants were developed using the PIG method, while stable GUS-expressing calli were multiplied during selection on MS medium containing 250 mg/l cefotaxime and 50 mg/l hygromycin, incubated in constant dark. A highly significant difference was observed between two methods of transformation for both frequency of GUS expression and GUS spots per callus. PIG-mediated transformation resulted in higher GUS expression compared to the Agrobacterium method. These results demonstrate that Dichanthium annulatum is amenable to Agrobacterium-mediated genetic transformation using a binary vector.