Response accuracy of ChatGPT 3.5 Copilot and Gemini in interpreting biochemical laboratory data a pilot study

With the release of ChatGPT at the end of 2022, a new era of thinking and technology use has begun. Artificial intelligence models (AIs) like Gemini (Bard), Copilot (Bing), and ChatGPT-3.5 have the potential to impact every aspect of our lives, including laboratory data interpretation. To assess the accuracy of ChatGPT-3.5, Copilot, and Gemini responses in evaluating biochemical data. Ten simulated patients' biochemical laboratory data, including serum urea, creatinine, glucose, cholesterol, triglycerides, low-density lipoprotein (LDL-c), and high-density lipoprotein (HDL-c), in addition to HbA1c, were interpreted by three AIs: Copilot, Gemini, and ChatGPT-3.5, followed by evaluation with three raters. The study was carried out using two approaches. The first encompassed all biochemical data. The second contained only kidney function data. The first approach indicated Copilot to have the highest level of accuracy, followed by Gemini and ChatGPT-3.5. Friedman and Dunn's post-hoc test revealed that Copilot had the highest mean rank; the pairwise comparisons revealed significant differences for Copilot vs. ChatGPT-3.5 (P = 0.002) and Gemini (P = 0.008). The second approach exhibited Copilot to have the highest accuracy of performance. The Friedman test with Dunn's post-hoc analysis showed Copilot to have the highest mean rank. The Wilcoxon Signed-Rank Test demonstrated an indistinguishable response (P = 0.5) of Copilot when all laboratory data were applied vs. the application of only kidney function data. Copilot is more accurate in interpreting biochemical data than Gemini and ChatGPT-3.5. Its consistent responses across different data subsets highlight its reliability in this context.

An Empirical Evaluation of Prompting Strategies for Large Language Models in Zero-Shot Clinical Natural Language Processing: Algorithm Development and Validation Study

BACKGROUND

Large language models (LLMs) have shown remarkable capabilities in natural language processing (NLP), especially in domains where labeled data are scarce or expensive, such as the clinical domain. However, to unlock the clinical knowledge hidden in these LLMs, we need to design effective prompts that can guide them to perform specific clinical NLP tasks without any task-specific training data. This is known as in-context learning, which is an art and science that requires understanding the strengths and weaknesses of different LLMs and prompt engineering approaches.

OBJECTIVE

The objective of this study is to assess the effectiveness of various prompt engineering techniques, including 2 newly introduced types-heuristic and ensemble prompts, for zero-shot and few-shot clinical information extraction using pretrained language models.

METHODS

This comprehensive experimental study evaluated different prompt types (simple prefix, simple cloze, chain of thought, anticipatory, heuristic, and ensemble) across 5 clinical NLP tasks: clinical sense disambiguation, biomedical evidence extraction, coreference resolution, medication status extraction, and medication attribute extraction. The performance of these prompts was assessed using 3 state-of-the-art language models: GPT-3.5 (OpenAI), Gemini (Google), and LLaMA-2 (Meta). The study contrasted zero-shot with few-shot prompting and explored the effectiveness of ensemble approaches.

RESULTS

The study revealed that task-specific prompt tailoring is vital for the high performance of LLMs for zero-shot clinical NLP. In clinical sense disambiguation, GPT-3.5 achieved an accuracy of 0.96 with heuristic prompts and 0.94 in biomedical evidence extraction. Heuristic prompts, alongside chain of thought prompts, were highly effective across tasks. Few-shot prompting improved performance in complex scenarios, and ensemble approaches capitalized on multiple prompt strengths. GPT-3.5 consistently outperformed Gemini and LLaMA-2 across tasks and prompt types.

CONCLUSIONS

This study provides a rigorous evaluation of prompt engineering methodologies and introduces innovative techniques for clinical information extraction, demonstrating the potential of in-context learning in the clinical domain. These findings offer clear guidelines for future prompt-based clinical NLP research, facilitating engagement by non-NLP experts in clinical NLP advancements. To the best of our knowledge, this is one of the first works on the empirical evaluation of different prompt engineering approaches for clinical NLP in this era of generative artificial intelligence, and we hope that it will inspire and inform future research in this area.

Research Progress in Structure Synthesis, Properties, and Applications of Small-Molecule Silicone Surfactants

Silicone surfactants have garnered significant research attention owing to their superior properties, such as wettability, ductility, and permeability. Small-molecular silicone surfactants with simple molecular structures outperform polymeric silicone surfactants in terms of surface activity, emulsification, wetting, foaming, and other areas. Moreover, silicone surfactants with small molecules exhibit a diverse and rich molecular structure. This review discusses various synthetic routes for the synthesis of different classes of surfactants, including single-chain, "umbrella" structure, double chain, bolaform, Gemini, and stimulus-responsive surfactants. The fundamental surface/interface properties of the synthesized surfactants are also highlighted. Additionally, these surfactants have demonstrated enormous potential in agricultural synergism, drug delivery, mineral flotation, enhanced oil recovery, separation, and extraction, and foam fire-fighting.

Effect of in situ mixed micellization of ester-functionalized gemini surfactant at different pHs on solubilization and cosolubilization of various polycyclic aromatic hydrocarbons of varying hydrophobicities

A pH controlled cleavability unfolds the 3-in-1 surfactant feature of an ester-bonded gemini surfactant, 2, 2'-[(oxybis (ethane-1,2-diyl))bis (oxy)]bis (N-hexadecyl-N,Ndimethyl-2-oxoethanaminium) dichloride (C16-C4O2-C16), by reinforcing in-situ mixed micellization between cleaved components at non-neutral pH (pH 3,12). The triplicity is assigned to two mixed-micelle variants at pH 3 and pH 12 besides the unhydrolyzed C16-C4O2-C16 at pH 7. The pH-controlled aggregation of such trichotomic surfactant dramatically enhances the micellar solubilization/cosolubilization of PAHs viz. naphthalene (Np), phenanthrene (Ph), pyrene (Py), perylene (Pe). The cosolubilization of binary/ternary PAH mixtures in such remarkable micellar assemblies at pH 3, 7 and 12 yields intriguing synergistic or antagonistic solubility outcomes correlated to PAH-PAH and PAH-micelle interactions. This study provides valuable insights into the potential applications of the ester-bonded gemini surfactant for the cosolubilization of undesirable hydrophobic compounds at natural sites having variable pH.

Red blood cell exchange for SARS-CoV-2: A Gemini of therapeutic opportunities

As of now, therapeutic strategies for the novel coronavirus (SARS-CoV-2) are limited and much focus has been placed on social distancing techniques to “flatten the curve”. Initial treatment efforts including ventilation and hydroxychloroquine garnered significant controversy and today, SARS-CoV-2 outbreaks are still occurring throughout the world. Needless to say, new therapeutic strategies are needed to combat this unprecedented pandemic. Nature Reviews Immunology recently published an article hypothesizing the pathogenesis of TAM (Tyro3, Axl, and Mer) receptor signaling in COVID-19. In it they expressed that hypercoagulation and immune hyper-reaction could occur secondary to decreased Protein S (PROS1). And hypoxia has been recently discovered to significantly decrease expression of PROS1. Regarding the cause of hypoxia in COVID-19; NIH funded research utilizing state-of-the-art topologies has recently demonstrated significant metabolomic, proteomic, and lipidomic structural aberrations in hemoglobin (Hb) secondary to infection with SARS-CoV-2. In this setting, Hb may be incapacitated and unable to respond to environmental variations, compromising RBCs and oxygen delivery to tissues. The use of red blood cell exchange would target hypoxia at its source; representing a Gemini of therapeutic opportunities.

Environmental behaviour and ecotoxicity of cationic surfactants towards marine organisms

Cationic surfactants are surface-active compounds that can be found in many products, including household and cleaning agents. As a consequence, they tend to be discarded into water streams, ultimately ending up in the aquatic environment. In spite of this environmental issue, studies describing their effects towards marine species are lacking. The aim of this study was therefore to evaluate the short-term exposure effects of two commercial cationic surfactants and three novel gemini surfactants on four marine species, the green microalgae Nannochloropsis gaditana and Tetraselmis chuii, the diatom Phaeodactylum tricornutum, and the crustacean Artemia salina. Furthermore, biodegradation and size distribution of the cationic surfactants in artificial seawater were also studied by UV-vis spectrophotometry and dynamic light scattering, respectively. Ecotoxicity tests revealed that the commercial cationic surfactant N-cetyl-N,N,N-trimethylammonium bromide is toxic to all tested marine species while N-dodecyl-N,N,N-trimethylammonium chloride and 1,4-bis-[N-(1-dodecyl)-N,N-dimethylammoniummethyl]benzene dibromide showed the lowest toxicity among the tested cationic surfactants. Besides the novel insights regarding the effects caused by these five cationic surfactants, this work opens prospects for the replacement of commercially available surfactants by more environmentally friendly alternatives.

Activity-Structure Study on the Peptide Fraction of AG2: a Potent In Vitro Transfection Agent

Gemini-based amphiphiles are candidates for biomedical applications. In fact, most of the gemini compounds described in the literature have been prepared to be used as new synthetic vectors in gene transfection. Our group carried out an activity-structure study starting from the structure of the gemini [AG2-C_18/]₂, which is an effective in vitro transfection reagent. We synthesized a series of novel amphiphilic amino acid derivatives of low molecular weight, named AGn-C_m (N), in which the same apolar region (m) of oleic or palmitic acid was maintained and the peptide region was modified by amino acid insertions, deletions, and substitutions. We also determined the transfection efficiency, critical micelle concentration, particle size, and ζ-potential for these derivatives. Amphiphiles AG10-C₁₆ and AG10-C₁₈ were more active at a lower N/P ratio than AG2-C₁₈. These amphiphiles showed no activity when lysine was replaced by ornithine, and the activity of all derivatives increased when there were more ornithine residues and a W/O = 1 ratio in the peptide region. It can be said that for AG10-C₁₆, these two structural requirements on the amino acid portion predominated over the type of aliphatic chain used.

The UVB1 Vitamin D analogue inhibits colorectal carcinoma progression

Vitamin D has been shown to display a wide variety of antitumour effects, but their therapeutic use is limited by its severe side effects. We have designed and synthesized a Gemini vitamin D analogue of calcitriol (UVB1) which has shown to display antineoplastic effects on different cancer cell lines without causing hypercalcemia. The aim of this work has been to investigate, by employing in silico, in vitro, and in vivo assays, whether UVB1 inhibits human colorectal carcinoma progression. We demonstrated that UVB1 induces apoptotic cell death and retards cellular migration and invasion of HCT116 colorectal carcinoma cells. Moreover, the analogue reduced the tumour volume in vivo, and modulated the expression of Bax, E-cadherin and nuclear β-catenin in tumour animal tissues without producing toxic effects. In silico analysis showed that UVB1 exhibits greater affinity for the ligand binding domain of vitamin D receptor than calcitriol, and that several characteristics in the three-dimensional conformation of VDR may influence the biological effects. These results demonstrate that the Gemini vitamin D analogue affects the growth of the colorectal cancer and suggest that UVB1 is a potential chemotherapeutic agent for treatment of this disease.

Use of surfactants for the remediation of contaminated soils: a review

Due to the great harm caused by soil contamination, there is an increasing interest to apply surfactants to the remediation of a variety of contaminated soils worldwide. This review article summarizes the findings of recent literatures regarding remediation of contaminated soils/sites using surfactants as an enhancing agent. For the surfactant-based remedial technologies, the adsorption behaviors of surfactants onto soil, the solubilizing capability of surfactants, and the toxicity and biocompatibility of surfactants are important considerations. Surfactants can enhance desorption of pollutants from soil, and promote bioremediation of organics by increasing bioavailability of pollutants. The removal of heavy metals and radionuclides from soils involves the mechanisms of dissolution, surfactant-associated complexation, and ionic exchange. In addition to the conventional ionic and nonionic surfactants, gemini surfactants and biosurfactants are also applied to soil remediation due to their benign features like lower critical micelle concentration (CMC) values and better biocompatibility. Mixed surfactant systems and combined use of surfactants with other additives are often adopted to improve the overall performance of soil washing solution for decontamination. Worldwide the field studies and full-scale remediation using surfactant-based technologies are yet limited, however, the already known cases reveal the good prospect of applying surfactant-based technologies to soil remediation.