United States

Despite significant advancements in pharmaceuticals and medical interventions, heart disease remains the leading cause of death worldwide. Although much research has been dedicated to understanding its risk factors, significant gaps persist, particularly in understudied populations. Addressing these gaps is essential to developing more effective, personalized strategies for preventing and treating heart disease. Current research often analyzes risk factors independently, without considering their potential interactions. As a result, treatments that benefit one patient may be ineffective for another, yet providers frequently adopt a one-size-fits-all approach. We can move toward a more personalized healthcare model by analyzing how risk factors, outcomes, diagnostics, and treatments vary between individuals. Understanding the complex interplay between genetic, social, and psychological determinants will help improve outcomes and bring us closer to truly individualized care on a global scale. This study aims to investigate factors that predict or protect against myocardial infarction (MI) in patients with a known risk factor of hyperlipidemia. Using regression modeling and subsets, the analysis will evaluate individual variables and their interactions. 


AMA Research Challenge 2025 

Risk Factors for Myocardial Infarction in Patients with Hyperlipidemia: A Logistical Regression Analysis

### Welcome!
The 62nd Annual Meeting of the Association for Computational Linguistics (ACL 2024) will take place in Bangkok, Thailand from August 11th to 16th, 2024. <br>Our Virtual Poster Sessions will take place online Thursday, August 22, 2024.<br><br><br>

You are required to register for this event. <br>**Please register [here](https://2024.aclweb.org/registration). **

If you have already registered, please check your inbox for an email from Underline granting you access to ACL 2024 content.


Please register!

ACL 2024

The 62nd Annual Meeting of the Association for Computational Linguistics (ACL 2024) will take place in Bangkok, Thailand from August 11th to 16th, 2024. More information will be announced soon.

poster

Congratulations to all the researchers who have qualified for the AMA Research Challenge virtual poster symposium and semifinals!

Welcome to the **largest national, multi-specialty research event** featuring research from nearly 1,000 medical students, residents, fellows, and international medical graduates across six topic areas - basic science, clinical and translational research, clinical vignettes, health systems science, medical education, and public health and health policy.

#### Access the Event
Use the buttons below or navigation menu to access the event. 

[![](https://assets.underline.io/markdown_image/1/image/d4076f2146939317a29eb9f69c53ad41.png)](https://underline.io/events/506/sessions?searchGroup=all)
[![](https://assets.underline.io/markdown_image/1/image/6bcee0b523a4874c4292ca0415d88dde.png)](https://underline.io/events/506/posters?searchGroup=all)

#### Event Details
The 2025 AMA Research Challenge includes the following symposium halls:
* **Semifinals hall:** Features research from 54 medical students, residents and international medical graduates that received top scores during first round scoring. 
* **Poster symposium hall:** Features research from nearly 1,000 medical students, residents and international medical graduates categorized by topics and subtopics.

**Here’s how to participate:**
* **View posters and presentations** on a variety of topics and specialties
* **Score posters in the semifinals** to help decide the top five finalists (AMA member exclusive opportunity)
* **Connect with presenters** and provide feedback to help advance their research
* **New this year: Watch live poster presentations** by topic area and engage directly with researchers

At the conclusion of this event, the AMA announced the top five finalists who advanced to the AMA Research Challenge final round, presented their research to an elite panel of judges, and competed to win a $10,000 grand prize presented by Laurel Road.

**Addison Shenk, a fourth-year medical student from Edward Via Virginia College of Osteopathic Medicine, was the winner of the 2025 AMA Research Challenge and was awarded the grand prize of $10,000, presented by Laurel Road.**

#### A message from Dr. Bobby Mukkamala, AMA President
<div style="position: relative; width: 100%; padding-bottom: 56.25%; height: 0; overflow: hidden;">
  <iframe 
    src="https://underline.io/embed/136894-ama-video?t=0"
    title="AMA Welcome Video"
    allowfullscreen
    style="position: absolute; top: 0; left: 0; width: 100%; height: 100%; border: 0;">
  </iframe>
</div>


#### Recognized Institutions
In 2025, 1,400 abstracts were submitted across six topic areas representing 174 medical schools and more than 150 residency and health care institutions. [See if your institution is represented this year!](https://drive.google.com/file/d/1eFrTqCmCeobx9pr5drJK2hllvsIpFNOY/view?usp=drive_link)

#### Grand Prize Sponsor
AMA would like to thank Laurel Road, a brand of KeyBank, for sponsoring a $10,000 grand prize for the winner of the AMA Research Challenge. Laurel Road and KeyBank are not involved in winner selection process. The winner will be announced during the AMA Research Challenge final round. Payment of the grand prize will be directly from Laurel Road/KeyBank. See the [AMA Research Challenge Official Rules for terms and conditions](https://www.ama-assn.org/system/files/research-challenge-official-rules.pdf).

[![](https://assets.underline.io/uploads/markdown_image/1/image/611e331d4a0eba0920f240cfc04b4d77.png)](https://www.laurelroad.com/)

#### Code of Conduct
<html>
AMA has a code of conduct for its meetings – we ask all attendees to adhere to this code to participate in this year’s meeting. Please read and <a href="https://www.ama-assn.org/about/code-conduct" target="_blank">learn more here</a> about your conduct requirements.
<br><br>
	</html>
	

#### Semifinals Judges

You need to log in with the email address you registered with. Access credentials have been sent to your email. 

Please be sure to check your spam and other email folders if you do not see an email confirmation right away.

Need help? Contact us at <support@underline.io>.

Please log in to explore this event.

If you have used the AMA Research Challenge platform previously, please click the ‘Log-in’ button below. If you do not remember your password, please click “forgot password” and follow prompts to reset your password. After resetting your password, please refresh your webpage to access the platform. 

Register to explore research in a variety of topics. AMA members get exclusive access to score posters in the semifinals. 

**Please use [this link here](https://www.ama-assn.org/member-benefits/events/ama-research-challenge) to register for free. **

Registration Is Required

Join the AMA Research Challenge poster symposium and semifinals to explore top research posters and score posters in the semifinals (an AMA member exclusive opportunity).

Background 
Environmental degradation due to industrial groundwater contamination is an emerging 
cardiovascular disease (CVD) risk factor. Our study examines the elevated CVD prevalence, 
including heart failure (HF) risk factors, in rural Punjab, India, near which a private distillery was 
linked to long-term groundwater contamination and environmental degradation. Our team 
primarily focused on CVD prevalence and its associations with environmental toxins. We also 
explored the role of socio-economic disparities on health outcomes. 
Methods 
A cross-sectional study of 4,885 residents in three villages adjacent to the distillery (Mahianwala 
Kala, Ratol Rohi, Mansoorwaal Kalan) was conducted from October 2022 to August 2023. 
Health camps with follow-up door-to-door surveys by trained medical personnel collected 
medical histories, environmental exposure details, and socio-demographic data. Key outcomes 
included both communicable and non-communicable diseases, such as hypertension, ischemic 
heart disease, heart failure, and cancer. Logistic regression assessed associations between 
CVD prevalence, socio-economic status, and pollution exposure, with groundwater testing for 
heavy metals and chemical oxygen demand (COD). 
Results 
The cohort demonstrated a high CVD burden, with 39% having hypertension (exceeding 
national averages by more than 50%), 34% having diagnosed CVD, and 21.3% of households 
reporting an immediate family history of myocardial infarction. Adjusted models highlighted 
elevated odds of heart disease (OR 1.42; 95% CI 1.31–1.51) and stroke (OR 1.29; 95% CI 
1.16–1.38), with higher clustering in low-income families. Concurrent systemic toxicity was also 
evident, as 60.5% reported dermatological conditions, 10.5% familial cancer, 20.8% with no 
history of smoking cited COPD-related dyspnea. We also observed elevated miscarriage rates 
and developmental disabilities in the studied population. Groundwater analysis confirmed 
contamination with cyanide (0.3 mg/L), lead (0.16 mg/L), and COD (320 mg/L), correlating with 
self-reported water discoloration and irregularities in taste. Families with polluted water access 
had a 1.5-fold increased CVD risk. 
Conclusion 
Industrial contamination was linked to higher cardiovascular morbidity and multi-organ health 
effects in our study's cohort in rural India. The link was exacerbated by socio-economic 
inequities. These findings highlight the role of environmental degradation in rise in CVD, 
particularly in low-resource settings.

Impact of Environmental Toxins and Socio-Economic Disparities on Cardiovascular Disease Burden: Insights from a Rural Indian Cohort Exposed to long term Industrial Pollution

Background: Retinal vascular occlusions are known causes of acute visual impairment. Prior studies have found an increased risk of retinal vascular occlusion among individuals with human immunodeficiency virus (HIV). However, these studies were conducted between 1983 and 1998, prior to the establishment of antiretroviral therapy (ART), which has since significantly improved HIV-related morbidity and mortality. While more recent case reports suggest a continued disease burden, no large-scale studies have examined this association in the ART era. We aim to determine the incidence and risk of retinal vascular occlusion, including retinal artery occlusion (RAO) and retinal vein occlusion (RVO), in HIV patients since widespread ART access.

Methods: We conducted a retrospective cohort analysis of healthcare data from May 2005 to May 2025 within the TriNetX global network. Adults ages 18 to 65 with HIV were compared to age-matched controls with dry eye syndrome. All patients were assessed by an ophthalmologist and excluded if they had prior history of retinal vascular occlusion. Propensity score matching (PSM) was used to balance demographics and comorbidities, including hypertension, diabetes mellitus, hyperlipidemia, and nicotine dependence. Incidence and relative risk (RR) of RAO and RVO were compared between cohorts. We calculated hazard ratios (HR) with 95% confidence intervals (CI) to stratify risk according to CD4 count (<100/µL vs. >100/µL) and viral load (< 200 copies/mL vs. > 200 copies/mL).

Results: After PSM, 26,736 patients were included in each cohort. HIV patients had an increased risk of RAO compared to controls (0.25% vs. 0.15%, RR=1.72, p<0.01); however, there was no significant difference in RVO outcomes (0.27% vs. 0.27%, RR=0.99, p=0.99). Among HIV patients, CD4 count <100/µL was associated with increased RAO risk (HR=1.83, p=0.02, 95% CI 1.11-2.99), but not RVO risk (HR=1.43, p=0.12, 95% CI 0.91-2.24). CD4 count >100/uL was not found to increase the risk of either RAO (HR=0.87, p=0.45, 95% CI 0.60-1.26) or RVO (HR=0.95, p=0.78, 95% CI 0.68-1.33). Viral load <200 copies/mL was not associated with a significant risk of RAO (HR=0.66, p=0.14, 95% CI 0.38-1.14) or RVO (HR=1.19, p=0.45, 95% CI 0.76-1.86).

Conclusion: HIV infection was associated with an increased risk of RAO, but not RVO, compared to controls. Given that the elevated RAO risk was concentrated among those with severe immunosuppression, these findings support early initiation of ART to mitigate HIV-associated RAO risk. Additionally, HIV screening may be warranted in atypical or unexplained cases of RAO.

Incidence and risk of retinal vascular occlusion in patients with HIV

Abstract Title: Risk Factors Influencing Axillofemoral Bypass Outcomes
Author(s): Niv Bhide, OMS II; Caila Tongco, OMS II; Hayden Gorham, OMS II; Christian Heck, PhD
Background: Axillofemoral bypass (AxFB) is an extra-anatomic procedure performed to restore blood flow to the lower limbs in cases of abdominal vascular blockage. AxFB is elected over the more invasive Aortofemoral Bypass (AoFB) when a patient is at greater perioperative risk due to comorbidities. This literature review aims to consolidate information on how managing risk factors affects post-operative outcomes of AxFB.
Methods: A literature review was conducted using PubMed, CINAHL, and ClinicalKey through Boolean searches with the following terms: ("axillofemoral bypass" OR "axillobifemoral bypass") AND ("graft surviv*" OR "mortality" OR "graft failure" OR "complications") AND ("risk factors") AND ("aortoiliac occlusive disease"). Inclusion criteria were full-text articles in English addressing graft patency. Case reports, case series, and articles on aortic aneurysms were excluded.
Results: A total of 50 articles were initially returned, 11 of which fit the exclusion/inclusion criteria. 3 articles provided hazard ratios or odds ratios comparing preoperative risk factors and postoperative outcomes. All 3 articles found worse outcomes (bypass occlusion, amputation, reintervention, cardiovascular events, and mortality) significantly associated with poor left ventricular function. 2 articles found worse outcomes with preoperative anticoagulant administration, anemia (Hb<10), and poor kidney function (ESKD or increased creatinine). Protective factors varied among the articles, but included postoperative anticoagulation, never-smokers, and those with a high BMI. Poor preoperative Society of Vascular Surgery (SVS) guideline adherence was noted, with 30% of patients not taking recommended statin therapy before surgery. 47-60% were current smokers, even though guidelines recommend smoking cessation before surgery.
Conclusion: Our results indicate a correlation between clinically manageable risk factors and worse outcomes. However, most of the current research on AxFB is focused on establishing the indications for AoFB vs AxFB, and many studies did not report complete data sets or multivariate analyses comparing comorbidities and outcomes. This review identifies the need for further data publication on how risk factor management and preoperative medications can impact bypass survival. This review also highlights an area for improvement of adherence to preoperative surgical guidelines.

Risk Factors Influencing Axillofemoral Bypass Outcomes

Background
Chronic orthopedic conditions such as arthritis often require specialist care to prevent long term complications. Children with arthritis have an increased risk of fracture, medication-induced complications and long-term emotional and social dysfunctions. However, according to the Arthritis Foundation, remission is possible in these children with early diagnosis and aggressive treatment. The objective of this study is to use National Survey of Children’s Health (NSCH) data to identify long-term implications due to difficulties accessing specialist care in children with arthritis.

Methods
This retrospective study utilized 2020-2021 NSCH data to evaluate whether children with arthritis had difficulties obtaining specialist care and the implications. All children with arthritis were stratified by demographic variables including age, sex and race. Multivariable logistic regression was used to compare children who faced challenges receiving specialist care with those who did not and the effect on their daily activities.

Results
A total of 113 children with arthritis who required access to specialty care were identified according to the study criteria. Most cases of arthritis presented in females (n= 73) with a median age of 15.1 years old. Out of the 113 children requiring specialty care, 37 of them had insufficient access to the needed care. Children with difficulty accessing specialist care were more likely to have interruptions in their daily activities with a Coefficient of 0.369 (95% Confidence Interval: 0.161 to 0.677; p=0.001).

Conclusion
This study faces limitations due to its small sample size. However, an overwhelming majority of the respondents that experienced difficulty receiving care for juvenile arthritis reported subsequent disruptions in daily activities. This underscores the need for future studies examining the barriers hindering certain children from obtaining specialist care. Understanding the root causes can facilitate the development of strategies to enhance specialist care accessibility, thereby improving outcomes for children with arthritis.

Health Equity and Childhood Arthritis: Barriers to Specialist Care and the Long-Term Impact on Daily Life

Background:
People experiencing homelessness face profound barriers to accessing palliative and end-of-life care, including stigma, fragmented services, and the lack of stable housing for hospice visits. As a result, many unhoused individuals die without comfort-focused care, social support, or dignity. Rocky Mountain Refuge (RMR) in Denver, Colorado, is one of only a handful of programs in the U.S. offering hospice support specifically for this population. By creating a safe space for those experiencing homelessness at the end of life, RMR fills a critical gap and offers lessons for advancing health equity in palliative care.

Methods:
Between February and October 2025, an internal evaluation of Rocky Mountain Refuge (RMR) was conducted using stakeholder surveys and qualitative interviews. The objective of this study was to evaluate RMR's unique care model to: a) inform program improvement and b) lay the foundation for a scalable model. Sixty stakeholders were invited to complete an online survey (24 responses, 40% response rate), and five qualitative interviews were conducted with supporters and donors. These participants were selected for their insight into RMR’s mission, operations, and sustainability. All qualitative data were analyzed using Braun and Clarke’s six-phase thematic analysis framework. 

Results:
Six key themes emerged: (1) Mission and dignity: RMR restores humanity and respect to those often excluded from end-of-life care; (2) Awareness and understanding: public misconceptions and discomfort surrounding death and homelessness hinder engagement; (3) Operational and financial challenges: limited capacity and funding instability constrain sustainability; (4) Policy and political context: funding and regulations create uncertainty in a shifting landscape; (5) Community trust and connection: personal relationships and in-person engagement foster understanding and support; and (6) Visibility and impact: storytelling, data sharing, and partnerships strengthen awareness and credibility.

Conclusions:
RMR represents an innovative, equity-driven model that restores dignity at the end of life for unhoused individuals. Findings highlight both the moral urgency and systemic challenges of sustaining such programs and underscore the importance of storytelling, community engagement, and policy navigation to advance equitable hospice and palliative care.

Dignity in Death: Evaluating a Unique End of Life Care Model that Bridges Hospice and Homelessness

Introduction 
Significant racial disparities in survival outcomes exist among Head and Neck Cancer (HNC) patients. Appropriate racial representation in clinical trials is essential to evaluate whether head and neck cancer (HNC) treatments are effective across diverse populations. Building upon a previous analysis by Zuckerman et al., that reported racial and ethnic disparity in HNC clinical trials, this current study assessed temporal changes in HNC clinical trial enrollment in the context of changing racial and ethnic disease prevalence between 2002 and 2020. 

Methods 
A comprehensive review of clinical trials focusing on HNC from January 2002 to December 2020 yielded 278 initial studies. After applying inclusion and exclusion criteria, 107 eligible trials were analyzed. Participant demographics were categorized into five standardized racial/ethnic groups aligned with the Surveillance, Epidemiology, and End Results (SEER) database classifications. Differences in racial group enrollment rates (as compared to their representation among HNC patients) were investigated using boxplots and scatterplots stratified by race and year, and with multiple linear regression.

Results 
Significant racial disparities in HNC clinical trials were identified, where White participants were overrepresented and minority groups, including Black (-0.121, 95% CI: [-0.145, -0.095]), Hispanic (-0.174, 95% CI: [-0.199, -0.149]), Asian/Pacific Islander (-0.129, 95% CI: [-0.155, -0.105]), and American Indian/Alaska Native (-0.104, 95% CI: [-0.128, -0.079]), were consistently underrepresented, with no change in representation over 18 years time period. This suggests that while previous analysis of this data had shown worsening of racial and ethnic diversity over time. This analysis examined these temporal changes in the context of changing disease prevalence, and illustrated stable enrollment proportions over time. 

Conclusion 
This work builds on work by Zuckerman et al. by adjusting for HNC patient demographics and by assessing annual changes in racial representation. This methodology revealed that there was no change in racial HNC trial representation over the last 20 years. It also reinforces the work by Zuckerman et al. by demonstrating that initiatives to ensure equitable representation are not working. In both studies, there was persistent over-representation of white HNC patients in HNC clinical trials. New approaches that facilitate equitable representation are needed to improve the generalizability and effectiveness of HNC treatment outcomes across all demographic groups.

Racial Disparities Between Head and Neck Cancer Clinical Trials and Head and Neck Cancer Patients in the United States: A Retrospective Study

Background 
Asthma is the most prevalent chronic illness in children worldwide, contributing to significant morbidity, healthcare utilization, and economic burden. In the United States, approximately 5 million children are currently affected. This review examines the environmental context and lifestyle determinants influencing pediatric asthma globally and within the U.S., with a special emphasis upon the Inland Empire (IE) in Southern California. The IE’s unique geographic landscape and role as a major transportation hub highlights its critical role for understanding how both environmental and structural factors exacerbate asthma burden in the IE pediatric population. 

Methods 
A narrative literature review and regional data analysis were conducted to identify environmental, socioeconomic, and healthcare accessibility factors behind pediatric asthma in the IE. PubMed, Scopus, and Google Scholar were searched for peer-reviewed articles published from 2000 to 2025. Data from the American Lung Association, CDC, and CDPH were also included. Key words included pediatric asthma, Inland Empire, air pollution, asthma disparity, emergency department utilization, hospitalizations, and socioeconomic status. 

Results 
Asthma has disproportionately affected the IE, consisting of San Bernardino (SB) and Riverside Counties. Geographically, the IE is surrounded by mountains that entrap pollutants such as vehicle emissions and industrial waste, limiting air circulation. Approximately 4 out of the top 5 most polluted cities in North America are located in the IE. Asthma ED visit rate in SB for children 0-17 years old was 60.5% and Riverside 59.3%, while the California average was 56.7%. Hospitalization rates for 0–4-year-olds were 24.4% for SB compared to California’s 17.3%. The elevated rates among school-aged children underscore the crucial need for interventions aimed at improving air quality, enhancing asthma management, and increasing access to preventive healthcare. 

Conclusion 
Pediatric asthma is a pressing matter of concern as it impairs the normal development of children, exerts a long-lasting impact on adolescent learning, disproportionately impacts underserved communities, and places a significant burden on the healthcare system. Here we highlight the unique determinants underlying the high prevalence of asthma in the 0–17-year-old population within the IE/SB, which includes heightened environmental risks (e.g. indoor allergens and air pollutants), socioeconomic barriers, lack of targeted public health initiatives. and unmet need for specialist-oriented care. We also emphasize the need for more research to improve the public health determinants of the IE. Future research should prioritize investigation of region-specific elements to develop tailored clinical strategies that will improve long-term outcomes for vulnerable pediatric patients in the IE.

Pediatric Asthma in Inland Empire: Analysis of Environmental Burden, Gaps in Care, and Unmet Needs

Artificial intelligence (AI) is transforming athletic training, offering novel tools to enhance performance, optimize recovery, and reduce injury risk. This systematic review synthesizes findings from 55 studies evaluating the integration of AI—particularly machine learning (ML) and deep learning (DL)—into sport-specific monitoring systems. These systems incorporate wearable data, video analysis, force measurements, and psychological profiling to support real-time decision-making in training and rehabilitation. Core applications are categorized across two main domains: (1) injury prevention, including movement analysis, fatigue/stimulus monitoring, and recovery forecasting; and (2) advancements to sport, including performance prediction, enhancement, and sport-specific adaptations. Across studies, models such as support vector machines (SVM), convolutional neural networks (CNN), long short-term memory (LSTM), and ensemble methods demonstrated predictive accuracies exceeding 90% in many contexts. This review highlights the potential of AI to augment the precision and scalability of athletic training, while also identifying current limitations, including data standardization, interpretability, and ethical oversight. Recommendations are offered to advance the responsible implementation of AI in athletic training for sport.


Artificial Intelligence in Athletic Training: A Systematic Review of Applications in Injury Prevention and Performance Optimization

Background: Asthma is a chronic lung disease characterized by airway hyperresponsiveness leading to reduction in quality of life. Volatile organic compounds (VOCs), easily evaporable chemicals found in fragrances, cleaning chemicals, and personal care products, are known to exacerbate asthma. The home can be a source of asthma triggers, including VOCs, leading to uncontrolled disease. Minimizing triggers in the home is essential for controlling asthma. Given that women are more likely to have greater severity of asthma than men, it is unknown whether there are gender differences in home VOC exposure. The purpose of this project is to examine any differences in VOC levels in the home by gender and education in older adults with asthma.

Methods: In this secondary analysis of a retrospective observational study, older adults ages 60 and above with asthma had VOC levels measured in their homes over a 24-hour period. Participants also completed baseline metrics of lung function and asthma control including forced expiratory volume (FEV-1), asthma control test (ACT), and Fractional Exhaled Nitric Oxide (FeNO). The 19 specific VOCs included in this analysis were detected indoors in 90% of homes or higher.

Results: Participants (N=184) were primarily White (77%), had a mean age at enrollment of 69±6.2 years, and included a female to male ratio of 2.7:1. Almost half (27.8%) were college educated, however 5.4% had less than a high school education. About one-third (33.7%) had an FEV1 percent predicted ≤79%, 36.4% had FeNO ≥25, and 56% had an ACT ≤19. The average of the total VOCs in each home was 68.36±87.37 parts per billion (ppb), the maximum was 603.69 ppb, minimum was 3.41 ppb, and the interquartile range (IQR) was 55.22 ppb. Acetone had the highest mean, median, and IQR of the 19 VOCs across the homes. Males had significantly higher levels of 124-trimethlybenzene in their homes (p= 0.005). Participants with less than a high school education had significantly higher levels of indoor freon113 and of carbon tetrachloride (p=0.019 and p= 0.013, respectively).

Conclusion: Differences in the concentration of specific VOCs in the homes of older adults with asthma by gender and education were identified. These findings may inform education and intervention strategies to reduce residential VOC exposure and improve asthma control in vulnerable populations.

Gender and education differences in residential levels of volatile organic compounds in older adults with asthma.

### Background
Multi-target peptide therapeutics targeting glucagon receptor (GCGR), glucagon-like peptide-1 receptor (GLP1R), and glucose-dependent insulinotropic polypeptide receptor (GIPR) represent a promising approach for treating diabetes and obesity. Triple agonist peptides demonstrate superior efficacy compared to single-target approaches, yet rational design remains computationally challenging due to complex sequence-structure-activity relationships. Existing methods, primarily based on convolutional neural networks, impose limitations including fixed sequence lengths and inadequate representation of molecular topology. Graph Attention Networks(GAT) offer advantages in capturing molecular structures and variable-length peptide sequences while providing interpretable insights into receptor-specific binding determinants.

### Objective
To develop and validate a GAT-based framework for designing novel triple agonist peptides with optimized binding affinity across GCGR, GLP1R, and GIPR, and to compare performance against convolutional neural network approaches.
Methods: A dataset of 234 peptide sequences with experimentally determined binding affinities was compiled from multiple sources. Peptides were represented as molecular graphs with seven-dimensional node features encoding physicochemical properties and positional information. The GAT architecture employed a shared encoder with task-specific prediction heads, implementing transfer learning to address limited GIPR training data. Performance was evaluated using 5-fold cross-validation and independent validation on 24 literature-derived sequences. A genetic algorithm framework was developed for peptide sequence optimization, incorporating multi-objective fitness evaluation based on predicted binding affinity, biological plausibility, and sequence novelty.

### Results
Cross-validation demonstrated robust GAT performance across all receptors, with GCGR achieving high accuracy (AUC-ROC: 0.915 ± 0.050), followed by GLP1R (AUC-ROC: 0.853 ± 0.059), and GIPR showing acceptable performance despite limited data (AUC-ROC: 0.907 ± 0.083). Comparative analysis revealed receptor-specific advantages: GAT significantly outperformed CNN for GCGR prediction (RMSE: 0.942 vs 1.209, p = 0.0013), while CNN maintained superior GLP1R performance (RMSE: 0.552 vs 0.723). Genetic algorithm optimization achieved substantial improvement over baseline, with 4.0% fitness enhancement and generation of 20 candidates exhibiting mean binding probabilities exceeding 0.5 across all targets.

### Conclusions
The GAT-based framework represents a significant advancement in computational peptide design, demonstrating receptor-specific advantages and robust optimization capabilities. Genetic algorithm optimization enables systematic exploration of sequence space while maintaining biological constraints. This approach provides a rational framework for prioritizing experimental validation efforts in triple agonist development.

Premium content

Downloads

Next from AMA Research Challenge 2025

Impact of Environmental Toxins and Socio-Economic Disparities on Cardiovascular Disease Burden: Insights from a Rural Indian Cohort Exposed to long term Industrial Pollution

Similar lecture

Automatic Extraction of Disease Risk Factors from Medical Publications