Predicting Current Glycated Hemoglobin Values in Adults: Development of an Algorithm From the Electronic Health Record

Identification of glutamine metabolism-related gene signature to predict colorectal cancer prognosis

Backgrounds: Colorectal cancer (CRC) is a highly malignant gastrointestinal malignancy with a poor prognosis, which imposes a significant burden on patients and healthcare providers globally. Previous studies have established that genes related to glutamine metabolism play a crucial role in the development of CRC. However, no studies have yet explored the prognostic significance of these genes in CRC. Methods: CRC patient data were downloaded from The Cancer Genome Atlas (TCGA), while glutamine metabolism-related genes were obtained from the Molecular Signatures Database (MSigDB) database. Univariate COX regression analysis and LASSO Cox regression were utilized to identify 15 glutamine metabolism-related genes associated with CRC prognosis. The risk scores were calculated and stratified into high-risk and low-risk groups based on the median risk score. The model's efficacy was assessed using Kaplan-Meier survival analysis and receiver operating characteristic (ROC) curve analysis. Cox regression analysis was employed to determine the risk score as an independent prognostic factor for CRC. Differential immune cell infiltration between the high-risk and low-risk groups was assessed using the ssGSEA method. The clinical applicability of the model was validated by constructing nomograms based on age, gender, clinical staging, and risk scores. Immunohistochemistry (IHC) was used to detect the expression levels of core genes. Results: We identified 15 genes related to glutamine metabolism in CRC: NLGN1, RIMKLB, UCN, CALB1, SYT4, WNT3A, NRCAM, LRFN4, PHGDH, GRM1, CBLN1, NRG1, GLYATL1, CBLN2, and VWC2. Compared to the high-risk group, the low-risk group demonstrated longer overall survival (OS) for CRC. Clinical correlation analysis revealed a positive correlation between the risk score and the clinical stage and TNM stage of CRC. Immune correlation analysis indicated a predominance of Th2 cells in the low-risk group. The nomogram exhibited excellent discriminatory ability for OS in CRC. Immunohistochemistry revealed that the core gene CBLN1 was expressed at a lower level in CRC, while GLYATL1 was expressed at a higher level. Conclusions: In summary, we have successfully identified and comprehensively analyzed a gene signature associated with glutamine metabolism in CRC for the first time. This gene signature consistently and reliably predicts the prognosis of CRC patients, indicating its potential as a metabolic target for individuals with CRC.

Exploring the potential link between MitoEVs and the immune microenvironment of periodontitis based on machine learning and bioinformatics methods

BACKGROUND

Periodontitis is a chronic inflammatory condition triggered by immune system malfunction. Mitochondrial extracellular vesicles (MitoEVs) are a group of highly heterogeneous extracellular vesicles (EVs) enriched in mitochondrial fractions. The objective of this research was to examine the correlation between MitoEVs and the immune microenvironment of periodontitis.

METHODS

Data from MitoCarta 3.0, GeneCards, and GEO databases were utilized to identify differentially expressed MitoEV-related genes (MERGs) and conduct functional enrichment and pathway analyses. The random forest and LASSO algorithms were employed to identify hub MERGs. Infiltration levels of immune cells in periodontitis and healthy groups were estimated using the CIBERSORT algorithm, and phenotypic subgroups of periodontitis based on hub MERG expression levels were explored using a consensus clustering method.

RESULTS

A total of 44 differentially expressed MERGs were identified. The random forest and LASSO algorithms identified 9 hub MERGs (BCL2L11, GLDC, CYP24A1, COQ2, MTPAP, NIPSNAP3A, FAM162A, MYO19, and NDUFS1). ROC curve analysis showed that the hub gene and logistic regression model presented excellent diagnostic and discriminating abilities. Immune infiltration and consensus clustering analysis indicated that hub MERGs were highly correlated with various types of immune cells, and there were significant differences in immune cells and hub MERGs among different periodontitis subtypes.

CONCLUSION

The periodontitis classification model based on MERGs shows excellent performance and can offer novel perspectives into the pathogenesis of periodontitis. The high correlation between MERGs and various immune cells and the significant differences between immune cells and MERGs in different periodontitis subtypes can clarify the regulatory roles of MitoEVs in the immune microenvironment of periodontitis. Future research should focus on elucidating the functional mechanisms of hub MERGs and exploring potential therapeutic interventions based on these findings.

Single-cell sequencing and transcriptome analyses in the construction of a liquid-liquid phase separation-associated gene model for rheumatoid arthritis

Background: Rheumatoid arthritis (RA) is a disabling autoimmune disease that affects multiple joints. Accumulating evidence suggests that imbalances in liquid-liquid phase separation (LLPS) can lead to altered spatiotemporal coordination of biomolecular condensates, which play important roles in carcinogenesis and inflammatory diseases. However, the role of LLPS in the development and progression of RA remains unclear. Methods: We screened RA and normal samples from GSE12021, GSE55235, and GSE55457 transcriptome datasets and GSE129087 and GSE109449 single-cell sequencing datasets from Gene Expression Omnibus database to investigate the pathogenesis of LLPS-related hub genes at the transcriptome and single cell sequencing levels. Machine learning algorithms and weighted gene co-expression network analysis were applied to screen hub genes, and hub genes were validated using correlation studies. Results: Differential analysis showed that 36 LLPS-related genes were significantly differentially expressed in RA, further random forest and support vector machine identified four and six LLPS-related genes, respectively, and weighted gene co-expression network analysis identified 396 modular genes. Hybridization of the three sets revealed two hub genes, MYC and MAP1LC3B, with AUCs of 0.907 and 0.911, respectively. Further ROC analysis of the hub genes in the GSE55457 dataset showed that the AUCs of MYC and MAP1LC3B were 0.815 and 0.785, respectively. qRT-PCR showed that the expression of MYC and MAP1LC3B in RA synovial tissues was significantly lower than that in the normal control synovial tissues. Correlation analysis between hub genes and the immune microenvironment and single-cell sequencing analysis revealed that both MYC and MAP1LC3B were significantly correlated with the degree of infiltration of various innate and acquired immune cells. Conclusion: Our study reveals a possible mechanism for LLPS in RA pathogenesis and suggests that MYC and MAP1LC3B may be potential novel molecular markers for RA with immunological significance.

Patient Perspectives on a Targeted Text Messaging Campaign to Encourage Screening for Diabetes: Qualitative Study

BACKGROUND

A sizeable proportion of prediabetes and diabetes cases among adults in the United States remain undiagnosed. Patient-facing clinical decision support (CDS) tools that leverage electronic health records (EHRs) have the potential to increase diabetes screening. Given the widespread mobile phone ownership across diverse groups, text messages present a viable mode for delivering alerts directly to patients. The use of unsolicited text messages to offer hemoglobin A1c (HbA1c) screening has not yet been studied. It is imperative to gauge perceptions of "cold texts" to ensure that information and language are optimized to promote engagement with text messages that affect follow-through with health behaviors.

OBJECTIVE

This study aims to gauge the perceptions of and receptiveness to text messages to inform content that would facilitate engagement with text messages intended to initiate a mobile health (mHealth) intervention for targeted screening. Messages were designed to invite those not already diagnosed with diabetes to make a decision to take part in HbA1c screening and walk them through the steps required to perform the behavior based solely on an automated text exchange.

METHODS

In total, 6 focus groups were conducted at Wake Forest Baptist Health (WFBH) between September 2019 and February 2020. The participants were adult patients without diabetes who had completed an in-person visit at the Family and Community Medicine Clinic within the previous year. We displayed a series of text messages and asked the participants to react to the message content and suggest improvements. Content was deductively coded with respect to the Health Belief Model (HBM) and inductively coded to identify other emergent themes that could potentially impact engagement with text messages.

RESULTS

Participants (N=36) were generally receptive to the idea of receiving a text-based alert for HbA1c screening. Plain language, personalization, and content, which highlighted perceived benefits over perceived susceptibility and perceived severity, were important to participants' understanding of and receptiveness to messages. The patient-physician relationship emerged as a recurring theme in which patients either had a desire or held an assumption that their provider would be working behind the scenes throughout each step of the process. Participants needed further clarification to understand the steps involved in following through with HbA1c screening and receiving results.

CONCLUSIONS

Our findings suggest that patients may be receptive to text messages that alert them to a risk of having an elevated HbA1c in direct-to-patient alerts that use cold texting. Using plain and positive language, integrating elements of personalization, and defining new processes clearly were identified by participants as modifiable content elements that could act as facilitators that would help overcome barriers to engagement with these messages. A patient's relationship with their provider and the financial costs associated with texts and screening may affect receptiveness and engagement in this process.

Development of a Gene Risk Signature for Patients of Pancreatic Cancer

BACKGROUND

Pancreatic cancer is a highly malignant solid tumor with a high lethality rate, but there is a lack of clinical biomarkers that can assess patient prognosis to optimize treatment.

METHODS

Gene-expression datasets of pancreatic cancer tissues and normal pancreatic tissues were obtained from the GEO database, and differentially expressed genes analysis and WGCNA analysis were performed after merging and normalizing the datasets. Univariate Cox regression analysis and Lasso Cox regression analysis were used to screen the prognosis-related genes in the modules with the strongest association with pancreatic cancer and construct risk signatures. The performance of the risk signature was subsequently validated by Kaplan-Meier curves, receiver operating characteristic (ROC), and univariate and multivariate Cox analyses.

RESULT

A three-gene risk signature containing CDKN2A, BRCA1, and UBL3 was established. Based on KM curves, ROC curves, and univariate and multivariate Cox regression analyses in the TRAIN cohort and TEST cohort, it was suggested that the three-gene risk signature had better performance in predicting overall survival.

CONCLUSION

This study identifies a three-gene risk signature, constructs a nomogram that can be used to predict pancreatic cancer prognosis, and identifies pathways that may be associated with pancreatic cancer prognosis.

Creating learning health systems and the emerging role of biomedical informatics

INTRODUCTION

The nature of information used in medicine has changed. In the past, we were limited to routine clinical data and published clinical trials. Today, we deal with massive, multiple data streams and easy access to new tests, ideas, and capabilities to process them. Whereas in the past getting information for decision-making was a challenge, now, it is how to analyze, evaluate and prioritize all that is readily available through the multitude of data-collecting devices. Clinicians must become adept with the tools needed to deal with the era of big data, requiring a major change in how we learn to make decisions. Major change is often met with resistance and questions about value. A Learning Health System is an enabler to encourage the development of such tools and demonstrate value in improved decision-making.

METHODS

We describe how we are developing a Biomedical Informatics program to help our medical institution's evolution as an academic Learning Health System, including strategy, training for house staff and examples of the role of informatics from operations to research.

RESULTS

We described an array of learning health system implementations and educational programs to improve healthcare and prepare a cadre of physicians with basic information technology skills. The programs have been well accepted with, for example, increasing interest and enrollment in the educational programs.

CONCLUSIONS

We are now in an era when large volumes of a wide variety of data are readily available. The challenge is not so much in the acquisition of data, but in assessing the quality, relevance and value of the data. The data we can get may not be the data we need. In the past, sources of data were limited, and trial results published in journals were the major source of evidence for decision making. The advent of powerful analytics systems has changed the concept of evidence. Clinicians will have to develop the skills necessary to work in the era of big data. It is not reasonable to expect that all clinicians will also be data scientists. However, understanding the role of AI and predictive analytics, and how to apply them, will become progressively more important. Programs such as the one being implemented at Wake Forest fill that need.

Improving Current Glycated Hemoglobin Prediction in Adults: Use of Machine Learning Algorithms With Electronic Health Records

Background

Predicting the risk of glycated hemoglobin (HbA_1c) elevation can help identify patients with the potential for developing serious chronic health problems, such as diabetes. Early preventive interventions based upon advanced predictive models using electronic health records data for identifying such patients can ultimately help provide better health outcomes.

Objective

Our study investigated the performance of predictive models to forecast HbA_1c elevation levels by employing several machine learning models. We also examined the use of patient electronic health record longitudinal data in the performance of the predictive models. Explainable methods were employed to interpret the decisions made by the black box models.

Methods

This study employed multiple logistic regression, random forest, support vector machine, and logistic regression models, as well as a deep learning model (multilayer perceptron) to classify patients with normal (<5.7%) and elevated (≥5.7%) levels of HbA_1c. We also integrated current visit data with historical (longitudinal) data from previous visits. Explainable machine learning methods were used to interrogate the models and provide an understanding of the reasons behind the decisions made by the models. All models were trained and tested using a large data set from Saudi Arabia with 18,844 unique patient records.

Results

The machine learning models achieved promising results for predicting current HbA_1c elevation risk. When coupled with longitudinal data, the machine learning models outperformed the multiple logistic regression model used in the comparative study. The multilayer perceptron model achieved an accuracy of 83.22% for the area under receiver operating characteristic curve when used with historical data. All models showed a close level of agreement on the contribution of random blood sugar and age variables with and without longitudinal data.

Conclusions

This study shows that machine learning models can provide promising results for the task of predicting current HbA_1c levels (≥5.7% or less). Using patients’ longitudinal data improved the performance and affected the relative importance for the predictors used. The models showed results that are consistent with comparable studies.

Identification of Vitamin D-related gene signature to predict colorectal cancer prognosis

Colorectal cancer (CRC) is one of the most common malignant carcinomas worldwide with poor prognosis, imposing an increasingly heavy burden on patients. Previous experiments and epidemiological studies have shown that vitamin D and vitamin D-related genes play a vital role in CRC. Therefore, we aimed to construct a vitamin D-related gene signature to predict prognosis in CRC. The CRC data from The Cancer Genome Atlas (TCGA) was performed as the training set. A total of 173 vitamin D-related genes in the TCGA CRC dataset were screened, and 17 genes associated with CRC prognosis were identified from them. Then, a vitamin D-related gene signature consisting of those 17 genes was established by univariate and multivariate Cox analyses. Moreover, four external datasets (GSE17536, GSE103479, GSE39582, and GSE17537) were used as testing set to validate the stability of this signature. The high-risk group presented a significantly poorer overall survival than low-risk group in both of training set and testing sets. Besides, the areas under the curve (AUCs) for signature on OS in training set at 1, 3, and 5 years were 0.710, 0.708, 0.710 respectively. The AUCs of the ROC curve in GSE17536 for 1, 3, and 5 years were 0.649, 0.654, and 0.694. These results indicated the vitamin D-related gene signature model could effectively predict the survival status of CRC patients. This vitamin D-related gene signature was also correlated with TNM stage in CRC clinical parameters, and the higher risk score from this model was companied with higher clinical stage. Furthermore, the high accuracy of this prognostic signature was validated and confirmed by nomogram model. In conclusion, we have proposed a novel vitamin D-related gene model to predict the prognosis of CRC, which will help provide new therapeutic targets and act as potential prognostic biomarkers for CRC.

Predicting Current Glycated Hemoglobin Levels in Adults From Electronic Health Records: Validation of Multiple Logistic Regression Algorithm

BACKGROUND

Electronic health record (EHR) systems generate large datasets that can significantly enrich the development of medical predictive models. Several attempts have been made to investigate the effect of glycated hemoglobin (HbA1c) elevation on the prediction of diabetes onset. However, there is still a need for validation of these models using EHR data collected from different populations.

OBJECTIVE

The aim of this study is to perform a replication study to validate, evaluate, and identify the strengths and weaknesses of replicating a predictive model that employed multiple logistic regression with EHR data to forecast the levels of HbA1c. The original study used data from a population in the United States and this differentiated replication used a population in Saudi Arabia.

METHODS

A total of 3 models were developed and compared with the model created in the original study. The models were trained and tested using a larger dataset from Saudi Arabia with 36,378 records. The 10-fold cross-validation approach was used for measuring the performance of the models.

RESULTS

Applying the method employed in the original study achieved an accuracy of 74% to 75% when using the dataset collected from Saudi Arabia, compared with 77% obtained from using the population from the United States. The results also show a different ranking of importance for the predictors between the original study and the replication. The order of importance for the predictors with our population, from the most to the least importance, is age, random blood sugar, estimated glomerular filtration rate, total cholesterol, non-high-density lipoprotein, and body mass index.

CONCLUSIONS

This replication study shows that direct use of the models (calculators) created using multiple logistic regression to predict the level of HbA1c may not be appropriate for all populations. This study reveals that the weighting of the predictors needs to be calibrated to the population used. However, the study does confirm that replicating the original study using a different population can help with predicting the levels of HbA1c by using the predictors that are routinely collected and stored in hospital EHR systems.

A Reinforcement Learning-Based Method for Management of Type 1 Diabetes: Exploratory Study

Background

Type 1 diabetes mellitus (T1DM) is characterized by chronic insulin deficiency and consequent hyperglycemia. Patients with T1DM require long-term exogenous insulin therapy to regulate blood glucose levels and prevent the long-term complications of the disease. Currently, there are no effective algorithms that consider the unique characteristics of T1DM patients to automatically recommend personalized insulin dosage levels.

Objective

The objective of this study was to develop and validate a general reinforcement learning (RL) framework for the personalized treatment of T1DM using clinical data.

Methods

This research presents a model-free data-driven RL algorithm, namely Q-learning, that recommends insulin doses to regulate the blood glucose level of a T1DM patient, considering his or her state defined by glycated hemoglobin (HbA_1c) levels, body mass index, engagement in physical activity, and alcohol usage. In this approach, the RL agent identifies the different states of the patient by exploring the patient’s responses when he or she is subjected to varying insulin doses. On the basis of the result of a treatment action at time step t, the RL agent receives a numeric reward, positive or negative. The reward is calculated as a function of the difference between the actual blood glucose level achieved in response to the insulin dose and the targeted HbA_1c level. The RL agent was trained on 10 years of clinical data of patients treated at the Mass General Hospital.

Results

A total of 87 patients were included in the training set. The mean age of these patients was 53 years, 59% (51/87) were male, 86% (75/87) were white, and 47% (41/87) were married. The performance of the RL agent was evaluated on 60 test cases. RL agent–recommended insulin dosage interval includes the actual dose prescribed by the physician in 53 out of 60 cases (53/60, 88%).

Conclusions

This exploratory study demonstrates that an RL algorithm can be used to recommend personalized insulin doses to achieve adequate glycemic control in patients with T1DM. However, further investigation in a larger sample of patients is needed to confirm these findings.