Comparison of Prompt Engineering and Fine-Tuning Strategies in Large Language Models in the Classification of Clinical Notes.

The emerging large language models (LLMs) are actively evaluated in various fields including healthcare. Most studies have focused on established benchmarks and standard parameters; however, the variation and impact of prompt engineering and fine-tuning strategies have not been fully explored. This study benchmarks GPT-3.5 Turbo, GPT-4, and Llama-7B against BERT models and medical fellows' annotations in identifying patients with metastatic cancer from discharge summaries. Results revealed that clear, concise prompts incorporating reasoning steps significantly enhanced performance. GPT-4 exhibited superior performance among all models. Notably, one-shot learning and fine-tuning provided no incremental benefit. The model's accuracy sustained even when keywords for metastatic cancer were removed or when half of the input tokens were randomly discarded. These findings underscore GPT-4's potential to substitute specialized models, such as PubMedBERT, through strategic prompt engineering, and suggest opportunities to improve open-source models, which are better suited to use in clinical settings.

Differential Transcriptomic Signatures of Small Airway Cell Cultures Derived from IPF and COVID-19-Induced Exacerbation of Interstitial Lung Disease.

Idiopathic pulmonary fibrosis (IPF) is a pathological condition wherein lung injury precipitates the deposition of scar tissue, ultimately leading to a decline in pulmonary function. Existing research indicates a notable exacerbation in the clinical prognosis of IPF patients following infection with COVID-19. This investigation employed bulk RNA-sequencing methodologies to describe the transcriptomic profiles of small airway cell cultures derived from IPF and post-COVID fibrosis patients. Differential gene expression analysis unveiled heightened activation of pathways associated with microtubule assembly and interferon signaling in IPF cell cultures. Conversely, post-COVID fibrosis cell cultures exhibited distinctive characteristics, including the upregulation of pathways linked to extracellular matrix remodeling, immune system response, and TGF-ß1 signaling. Notably, BMP signaling levels were elevated in cell cultures derived from IPF patients compared to non-IPF control and post-COVID fibrosis samples. These findings underscore the molecular distinctions between IPF and post-COVID fibrosis, particularly in the context of signaling pathways associated with each condition. A better understanding of the underlying molecular mechanisms holds the promise of identifying potential therapeutic targets for future interventions in these diseases.

CRISPRi-TnSeq: A genome-wide high-throughput tool for bacterial essential-nonessential genetic interaction mapping.

Genetic interaction networks can help identify functional connections between genes and pathways, which can be leveraged to establish (new) gene function, drug targets, and fill pathway gaps. Since there is no optimal tool that can map genetic interactions across many different bacterial strains and species, we develop CRISPRi-TnSeq, a genome-wide tool that maps genetic interactions between essential genes and nonessential genes through the knockdown of a targeted essential gene (CRISPRi) and the simultaneous knockout of individual nonessential genes (Tn-Seq). CRISPRi-TnSeq thereby identifies, on a genome-wide scale, synthetic and suppressor-type relationships between essential and nonessential genes, enabling the construction of essential-nonessential genetic interaction networks. To develop and optimize CRISPRi-TnSeq, CRISPRi strains were obtained for 13 essential genes in Streptococcus pneumoniae, involved in different biological processes including metabolism, DNA replication, transcription, cell division and cell envelope synthesis. Transposon-mutant libraries were constructed in each strain enabling screening of ∼24,000 gene-gene pairs, which led to the identification of 1,334 genetic interactions, including 754 negative and 580 positive genetic interactions. Through extensive network analyses and validation experiments we identify a set of 17 pleiotropic genes, of which a subset tentatively functions as genetic capacitors, dampening phenotypic outcomes and protecting against perturbations. Furthermore, we focus on the relationships between cell wall synthesis, integrity and cell division and highlight: 1) how essential gene knockdown can be compensated by rerouting flux through nonessential genes in a pathway; 2) the existence of a delicate balance between Z-ring formation and localization, and septal and peripheral peptidoglycan (PG) synthesis to successfully accomplish cell division; 3) the control of c-di-AMP over intracellular K + and turgor, and thereby modulation of the cell wall synthesis machinery; 4) the dynamic nature of cell wall protein CozEb and its effect on PG synthesis, cell shape morphology and envelope integrity; 5) functional dependency between chromosome decatenation and segregation, and the critical link with cell division, and cell wall synthesis. Overall, we show that CRISPRi-TnSeq uncovers genetic interactions between closely functionally linked genes and pathways, as well as disparate genes and pathways, highlighting pathway dependencies and valuable leads for gene function. Importantly, since both CRISPRi and Tn-Seq are widely used tools, CRISPRi-TnSeq should be relatively easy to implement to construct genetic interaction networks across many different microbial strains and species.

Sample Preparation and Differential Gene Expression Analysis of Human Cancer Cell Lines by RNA Sequencing.

RNA sequencing (RNA-seq) is a method used for the high-throughput quantification of mRNA in a biological sample. It is widely used to investigate differential gene expression between drug-resistant and sensitive cancers to identify genetic mediators of drug resistance. Here, we describe a comprehensive experimental and bioinformatic pipeline to isolate mRNA from human cell lines, prepare mRNA libraries for next-generation sequencing, and perform post-sequencing bioinformatics analyses.

A genome-wide atlas of antibiotic susceptibility targets and pathways to tolerance.

Detailed knowledge on how bacteria evade antibiotics and eventually develop resistance could open avenues for novel therapeutics and diagnostics. It is thereby key to develop a comprehensive genome-wide understanding of how bacteria process antibiotic stress, and how modulation of the involved processes affects their ability to overcome said stress. Here we undertake a comprehensive genetic analysis of how the human pathogen Streptococcus pneumoniae responds to 20 antibiotics. We build a genome-wide atlas of drug susceptibility determinants and generated a genetic interaction network that connects cellular processes and genes of unknown function, which we show can be used as therapeutic targets. Pathway analysis reveals a genome-wide atlas of cellular processes that can make a bacterium less susceptible, and often tolerant, in an antibiotic specific manner. Importantly, modulation of these processes confers fitness benefits during active infections under antibiotic selection. Moreover, screening of sequenced clinical isolates demonstrates that mutations in genes that decrease antibiotic sensitivity and increase tolerance readily evolve and are frequently associated with resistant strains, indicating such mutations could be harbingers for the emergence of antibiotic resistance.

A genetic model of methionine restriction extends Drosophila health- and lifespan.

Loss of metabolic homeostasis is a hallmark of aging and is characterized by dramatic metabolic reprogramming. To analyze how the fate of labeled methionine is altered during aging, we applied 13C5-Methionine labeling to Drosophila and demonstrated significant changes in the activity of different branches of the methionine metabolism as flies age. We further tested whether targeted degradation of methionine metabolism components would "reset" methionine metabolism flux and extend the fly lifespan. Specifically, we created transgenic flies with inducible expression of Methioninase, a bacterial enzyme capable of degrading methionine and revealed methionine requirements for normal maintenance of lifespan. We also demonstrated that microbiota-derived methionine is an alternative and important source in addition to food-derived methionine. In this genetic model of methionine restriction (MetR), we also demonstrate that either whole-body or tissue-specific Methioninase expression can dramatically extend Drosophila health- and lifespan and exerts physiological effects associated with MetR. Interestingly, while previous dietary MetR extended lifespan in flies only in low amino acid conditions, MetR from Methioninase expression extends lifespan independently of amino acid levels in the food. Finally, because impairment of the methionine metabolism has been previously associated with the development of Alzheimer's disease, we compared methionine metabolism reprogramming between aging flies and a Drosophila model relevant to Alzheimer's disease, and found that overexpression of human Tau caused methionine metabolism flux reprogramming similar to the changes found in aged flies. Altogether, our study highlights Methioninase as a potential agent for health- and lifespan extension.

Downregulation of the tyrosine degradation pathway extends Drosophila lifespan.

Aging is characterized by extensive metabolic reprogramming. To identify metabolic pathways associated with aging, we analyzed age-dependent changes in the metabolomes of long-lived Drosophila melanogaster. Among the metabolites that changed, levels of tyrosine were increased with age in long-lived flies. We demonstrate that the levels of enzymes in the tyrosine degradation pathway increase with age in wild-type flies. Whole-body and neuronal-specific downregulation of enzymes in the tyrosine degradation pathway significantly extends Drosophila lifespan, causes alterations of metabolites associated with increased lifespan, and upregulates the levels of tyrosine-derived neuromediators. Moreover, feeding wild-type flies with tyrosine increased their lifespan. Mechanistically, we show that suppression of ETC complex I drives the upregulation of enzymes in the tyrosine degradation pathway, an effect that can be rescued by tigecycline, an FDA-approved drug that specifically suppresses mitochondrial translation. In addition, tyrosine supplementation partially rescued lifespan of flies with ETC complex I suppression. Altogether, our study highlights the tyrosine degradation pathway as a regulator of longevity.