Mutational Analysis of Intragenic and Intergenic Variants of Uterine Cancer and Super Enhancer Associations
Abstract
Mutational variants and changes in the genetic code serve as the foundational cause for many diseases and syndromes. Amongst the many diseases, cancer has been displayed to be often consequential due to mutational variants. However, the progression and penetrance of phenotypes consequential to uterine cancers are not well understood. This study focuses on addressing this through mutational analysis of uterine cancer data acquired from the NIH TCGA database and further evaluating the potential association between associated mutations and the relative theoretical and observed consequences through modelling and existing data [1]
Weinstein JN, Collisson EA, Mills GB, Shaw KR, Ozenberger BA, Ellrott K, Shmulevich I, Sander C, Stuart JM. The cancer genome atlas pan-cancer analysis project. Nature genetics. 2013 Oct;45(10):1113-20.
Evaluating the Efficiency of 2A Sequences Due to Gene Size Variability in Polycistronic Vectors
Abstract
Multi-peptide producing processes have been an emerging topic. Originating from viruses, two primary modalities of multi-peptide-producing vectors have been observed: IRES sequences and 2A sequences. The latter is of great interest due to the short sequence length, diversity, and plasmid transfection efficiency, and has been extensively studied to identify various modifiers to improve cleavage efficiency[1]. Furthermore, various combinations of 2A sequences have been observed to play different cleavage efficiencies when used together, posing a challenge in effective, functional polycistronic vector disease. Additionally, little is known regarding the effects of gene characteristics in effects on 2A peptide efficiency. This study focuses on identifying various gene characteristics and applying them via vector design to observe the differences in cleavage efficiency amongst various 2A peptides. This study will provide greater insight into the effects of gene content when designing polycistronic vectors and the potential application in transcript synthesis and relevant applications.
References:
Wang X, Marchisio MA. Synthetic polycistronic sequences in eukaryotes. Synthetic and Systems Biotechnology. 2021 Dec 1;6(4):254-61.
Liu Z, Chen O, Wall JB, Zheng M, Zhou Y, Wang L, Ruth Vaseghi H, Qian L, Liu J. Systematic comparison of 2A peptides for cloning multi-genes in a polycistronic vector. Scientific reports. 2017 May 19;7(1):2193.
Evaluating the Potential Application of Accessible LLMs in Simulation-based Patient-Physician Cases and Development of Virtual Patient Cases for Medical-Based Education
Abstract
AI models have become extremely popular as prospective tools for healthcare and potential involvement in medical interactions, and eventually diagnosis. Large language models (LLM), such as ChatGPT, are amongst the most popular, with a diverse set of functions and roles, including involvement in healthcare and medical education. Companies such as Google have begun implementation of medical diagnostic AI models such as AMIE (Articulate Medical Intelligence Performer), which has shown promising results to advance both aspects of diagnostic capabilities and patient-physician interactions. Other LLMs have been developed that attempt to target specific aspects of patient-physician relationship, such as HAILEY, an AI chatbot aimed at improving empathy for text-based patient-physician interactions.
Another area where LLM and machine learning can assist is through simulation-based and conversational learning for medical education. With progressive advancement in LLM and machine learning, there is great potential in applying these models towards medical teaching through implementation of virtual standardized patient (VSP) cases. Although the aforementioned AMIE infrastructure is being developed, current simulation-based practice may be utilized through other LLMs, such have been put into question their legitimacy in appropriate emotional and objective responses in patient-physician interactions. In this study, we look and observe how various LLMs develop and present clinical scenarios based on presentation to various specialties stipulated by a role-play between a VSP and a specialist. Furthermore, complex emotional factors, such as stated tonality associated with emotions that could be considered to be involved with a given specialty, will be taken into consideration. The scenario responses and patient-based responses will be applied into the current framework amongst various LLMs and observed for relative changes in empathetic responses and medical accuracy. The findings of this study will enhance our understanding of LLM and the potential for currently available LLM models to be applied in creating clinical scenarios, relative biases developed by specialities perceived within LLMs, as well as cautions to consider regarding application in medical education.
References:
1.Scherr R, Halaseh FF, Spina A, Andalib S, Rivera R. ChatGPT interactive medical simulations for early clinical education: case study. JMIR Medical Education. 2023 Nov 10;9:e49877.
2. McDuff D, Schaekermann M, Tu T, Palepu A, Wang A, Garrison J, Singhal K, Sharma Y, Azizi S, Kulkarni K, Hou L. Towards accurate differential diagnosis with large language models. arXiv preprint arXiv:2312.00164. 2023 Nov 30.
3. Sharma A, Lin IW, Miner AS, Atkins DC, Althoff T. Human–AI collaboration enables more empathic conversations in text-based peer-to-peer mental health support. Nature Machine Intelligence. 2023 Jan;5(1):46-57.
4.Gray M, Baird A, Sawyer T, James J, DeBroux T, Bartlett M, Krick J, Umoren R. Increasing Realism and Variety of Virtual Patient Dialogues for Prenatal Counseling Education Through a Novel Application of ChatGPT: Exploratory Observational Study. JMIR Medical Education. 2024 Feb 1;10:e50705.