Distilling large language models for matching patients to clinical trials.

OBJECTIVE: The objective of this study is to systematically examine the efficacy of both proprietary (GPT-3.5, GPT-4) and open-source large language models (LLMs) (LLAMA 7B, 13B, 70B) in the context of matching patients to clinical trials in healthcare. MATERIALS AND METHODS: The study employs a multifaceted evaluation framework, incorporating extensive automated and human-centric assessments along with a detailed error analysis for each model, and assesses LLMs' capabilities in analyzing patient eligibility against clinical trial's inclusion and exclusion criteria. To improve the adaptability of open-source LLMs, a specialized synthetic dataset was created using GPT-4, facilitating effective fine-tuning under constrained data conditions. RESULTS: The findings indicate that open-source LLMs, when fine-tuned on this limited and synthetic dataset, achieve performance parity with their proprietary counterparts, such as GPT-3.5. DISCUSSION: This study highlights the recent success of LLMs in the high-stakes domain of healthcare, specifically in patient-trial matching. The research demonstrates the potential of open-source models to match the performance of proprietary models when fine-tuned appropriately, addressing challenges like cost, privacy, and reproducibility concerns associated with closed-source proprietary LLMs. CONCLUSION: The study underscores the opportunity for open-source LLMs in patient-trial matching. To encourage further research and applications in this field, the annotated evaluation dataset and the fine-tuned LLM, Trial-LLAMA, are released for public use.

Suicidal Behaviour, including Ideation and Self-Harm, in Young Migrants: A Systematic Review.

Young people experience high rates of suicidal ideation, self-harm, suicide attempt and death due to suicide. As a result of increasing globalisation, young people are increasingly mobile and can migrate from one country to another seeking educational and employment opportunities. With a growing number of young migrants, it is important to understand the prevalence of suicidal behaviour among this population group. We systematically searched Medline, Embase, and PsycINFO from inception until 31 March 2022. Eligible studies were those providing data on suicidal ideation, self-harm, suicide attempt, and death due to suicide. Seventeen studies were included in the review, some of which provided data on multiple outcomes of interest. Twelve studies provided data on suicidal ideation, five provided data on self-harm, eight provided data on suicide attempt, and one study had data on suicide death among young migrants. The quality of the included studies was varied and limited. The studies included in this review commonly reported that young migrants experience higher rates of self-harm and suicide attempt, but no major differences in suicidal ideation and suicide death compared to non-migrant young people. However, the limited number of studies focused on suicidal behaviour among young migrants highlights the need for further high-quality studies to capture accurate information. This will enable the development of policies and interventions that reduce the risk of suicidal behaviour among young migrants.

Engineered Humicola insolens cutinase for efficient cellulose acetate deacetylation.

Cutinases comprise a family of esterases with broad hydrolytic activity for chain and pendant ester groups. This work aimed to identify and improve an efficient cutinase for cellulose acetate (CA) deacetylation. The development of a mild method for CA fiber surface deacetylation will result in improved surface hydrophilicity and reactivity while, when combined with cellulases, a route to the full recycling of CA to acetate and glucose. In this study, the comparative CA deacetylation activity of four homologous wild-type (wt) fungal cutinases from Aspergillus oryzae (AoC), Thiellavia terrestris (TtC), Fusarium solani (FsC), and Humicola insolens (HiC) was determined by analysis of CA deacetylation kinetics. wt-HiC had the highest catalytic efficiency (≈32 [cm2 L-1 ]-1 h-1 ). Comparison of wt-cutinase catalytic constants revealed that differences in catalytic efficiency are primarily due to corresponding variations in corresponding substrate binding constants. Docking studies with model tetrameric substrates also revealed structural origins for differential substrate binding amongst these cutinases. Comparative docking studies of HiC point mutations led to the identification of two important rationales for engineering cutinases for CA deacetylation: (i) create a tight but not too closed binding groove, (ii) allow for hydrogen bonding in the extended region around the active site. Rationally designed HiC with amino acid substitutions I36S, predicted to hydrogen bond to CA, combined with F70A, predicted to remove steric constraints, showed a two-fold improvement in catalytic efficiency. Continued cutinase optimization guided by a detailed understanding of structure-activity relationships, as demonstrated here, will be an important tool to developing practical cutinases for commercial green chemistry technologies.

Purification and characterization of benzyl alcohol- and benzaldehyde- dehydrogenase from Pseudomonas putida CSV86.

Pseudomonas putida CSV86 utilizes benzyl alcohol via catechol and methylnaphthalenes through detoxification pathway via hydroxymethylnaphthalenes and naphthaldehydes. Based on metabolic studies, benzyl alcohol dehydrogenase (BADH) and benzaldehyde dehydrogenase (BZDH) were hypothesized to be involved in the detoxification pathway. BADH and BZDH were purified to apparent homogeneity and were (1) homodimers with subunit molecular mass of 38 and 57 kDa, respectively, (2) NAD(+) dependent, (3) broad substrate specific accepting mono- and di-aromatic alcohols and aldehydes but not aliphatic compounds, and (4) BADH contained iron and magnesium, while BZDH contained magnesium. BADH in the forward reaction converted alcohol to aldehyde and required NAD(+), while in the reverse reaction it reduced aldehyde to alcohol in NADH-dependent manner. BZDH showed low K (m) value for benzaldehyde as compared to BADH reverse reaction. Chemical cross-linking studies revealed that BADH and BZDH do not form multi-enzyme complex. Thus, the conversion of aromatic alcohol to acid is due to low K (m) and high catalytic efficiency of BZDH. Phylogenetic analysis revealed that BADH is a novel enzyme and diverged during the evolution to gain the ability to utilize mono- and di-aromatic compounds. The wide substrate specificity of these enzymes enables strain to detoxify methylnaphthalenes to naphthoic acids efficiently.

Sequential utilization of substrates by Pseudomonas putida CSV86: signatures of intermediate metabolites and online measurements.

Pseudomonas putida CSV86 preferentially utilizes aromatics over glucose and co-metabolizes them with organic acids. On aromatics plus glucose, CSV86 utilized aromatics first with concomitant appearance of transient metabolites such as salicylate, benzaldehyde and benzoate. Citrate was the main extracellular metabolite observed during glucose uptake. The strain showed simultaneous utilization of organic acids and aromatic compounds. Based on the metabolite analysis and growth profiles, we hypothesize that the repression of glucose utilization could be due to organic acid intermediates generated from aromatic compound metabolism. The online measurements indicate the instantaneous metabolic state of the culture. For example, the CO(2) evolution and agitation speed show peak values during the two growth phases in the diauxic growth while dissolved oxygen values show decrease at the corresponding durations. These measurements correlated well with the offline measurements but provided a better time resolution of the process.

Conjugative transfer of preferential utilization of aromatic compounds from Pseudomonas putida CSV86.

Pseudomonas putida CSV86 utilizes naphthalene (Nap), salicylate (Sal), benzyl alcohol (Balc), and methylnaphthalene (MN) preferentially over glucose. Methylnaphthalene is metabolized by ring-hydroxylation as well as side-chain hydroxylation pathway. Although the degradation property was found to be stable, the frequency of obtaining Nap(-)Sal(-)MN(-)Balc(-) phenotype increased to 11% in the presence of curing agents. This property was transferred by conjugation to Stenotrophomonas maltophilia CSV89 with a frequency of 7 x 10(-8) per donor cells. Transconjugants were Nap(+)Sal(+)MN(+)Balc(+) and metabolized MN by ring- as well as side-chain hydroxylation pathway. Transconjugants also showed the preferential utilization of aromatic compounds over glucose indicating transfer of the preferential degradation property. The transferred properties were lost completely when transconjugants were grown on glucose or 2YT. Attempts to detect and isolate plasmid DNA from CSV86 and transconjugants were unsuccessful. Transfer of degradation genes and its subsequent loss from the transconjugants was confirmed by PCR using primers specific for 1,2-dihydroxynaphthalene dioxygenase and catechol 2,3-dioxygenase (C23O) as well as by DNA-DNA hybridizations using total DNA as template and C23O PCR fragment as a probe. These results indicate the involvement of a probable conjugative element in the: (i) metabolism of aromatic compounds, (ii) ring- and side-chain hydroxylation pathways for MN, and (iii) preferential utilization of aromatics over glucose.

Modulation of glucose transport causes preferential utilization of aromatic compounds in Pseudomonas putida CSV86.

Pseudomonas putida CSV86 utilizes aromatic compounds in preference to glucose and coutilizes aromatics and organic acids. Protein analysis of cells grown on different carbon sources, either alone or in combination, revealed that a 43-kDa periplasmic-space protein was induced by glucose and repressed by aromatics and succinate. Two-dimensional gel electrophoresis and liquid chromatography-tandem mass spectrometry analysis identified this protein as closely resembling the sugar ABC transporter of Pseudomonas putida KT2440. A partially purified 43-kDa protein showed glucose binding activity and was specific for glucose. The results demonstrate that the aromatic- and organic acid-mediated repression of a periplasmic-space glucose binding protein and consequent inhibition of glucose transport are responsible for this strain's ability to utilize aromatics and organic acids in preference to glucose.

Metabolic diversity in bacterial degradation of aromatic compounds.

Aromatic compounds pose a major threat to the environment, being mutagenic, carcinogenic, and recalcitrant. Microbes, however, have evolved the ability to utilize these highly reduced and recalcitrant compounds as a potential source of carbon and energy. Aerobic degradation of aromatics is initiated by oxidizing the aromatic ring, making them more susceptible to cleavage by ring-cleaving dioxygenases. A preponderance of aromatic degradation genes on plasmids, transposons, and integrative genetic elements (and their shuffling through horizontal gene transfer) have lead to the evolution of novel aromatic degradative pathways. This enables the microorganisms to utilize a multitude of aromatics via common routes of degradation leading to metabolic diversity. In this review, we emphasize the exquisiteness and relevance of bacterial degradation of aromatics, interlinked degradative pathways, genetic and metabolic regulation, carbon source preference, and biosurfactant production. We have also explored the avenue of metagenomics, which opens doors to a plethora of uncultured and uncharted microbial genetics and metabolism that can be used effectively for bioremediation.

Purification and characterization of 1-naphthol-2-hydroxylase from carbaryl-degrading Pseudomonas strain c4.

Pseudomonas sp. strain C4 metabolizes carbaryl (1-naphthyl-N-methylcarbamate) as the sole source of carbon and energy via 1-naphthol, 1,2-dihydroxynaphthalene, and gentisate. 1-Naphthol-2-hydroxylase (1-NH) was purified 9.1-fold to homogeneity from Pseudomonas sp. strain C4. Gel filtration and sodium dodecyl sulfate-polyacrylamide gel electrophoresis showed that the enzyme is a homodimer with a native molecular mass of 130 kDa and a subunit molecular mass of 66 kDa. The enzyme was yellow, with absorption maxima at 274, 375, and 445 nm, indicating a flavoprotein. High-performance liquid chromatography analysis of the flavin moiety extracted from 1-NH suggested the presence of flavin adenine dinucleotide (FAD). Based on the spectral properties and the molar extinction coefficient, it was determined that the enzyme contained 1.07 mol of FAD per mol of enzyme. Although the enzyme accepts electrons from NADH, it showed maximum activity with NADPH and had a pH optimum of 8.0. The kinetic constants K(m) and V(max) for 1-naphthol and NADPH were determined to be 9.6 and 34.2 microM and 9.5 and 5.1 micromol min(-1) mg(-1), respectively. At a higher concentration of 1-naphthol, the enzyme showed less activity, indicating substrate inhibition. The K(i) for 1-naphthol was determined to be 79.8 microM. The enzyme showed maximum activity with 1-naphthol compared to 4-chloro-1-naphthol (62%) and 5-amino-1-naphthol (54%). However, it failed to act on 2-naphthol, substituted naphthalenes, and phenol derivatives. The enzyme utilized one mole of oxygen per mole of NADPH. Thin-layer chromatographic analysis showed the conversion of 1-naphthol to 1,2-dihydroxynaphthalene under aerobic conditions, but under anaerobic conditions, the enzyme failed to hydroxylate 1-naphthol. These results suggest that 1-NH belongs to the FAD-containing external flavin mono-oxygenase group of the oxidoreductase class of proteins.

Inducible uptake and metabolism of glucose by the phosphorylative pathway in Pseudomonas putida CSV86.

Pseudomonas putida CSV86 utilizes glucose, naphthalene, methylnaphthalene, benzyl alcohol and benzoate as the sole source of carbon and energy. Compared with glucose, cells grew faster on aromatic compounds as well as on organic acids. The organism failed to grow on gluconate, 2-ketogluconate, fructose and mannitol. Whole-cell oxygen uptake, enzyme activity and metabolic studies suggest that in strain CSV86 glucose utilization is exclusively by the intracellular phosphorylative pathway, while in Stenotrophomonas maltophilia CSV89 and P. putida KT2442 glucose is metabolized by both direct oxidative and indirect phosphorylative pathways. Cells grown on glucose showed five- to sixfold higher activity of glucose-6-phosphate dehydrogenase compared with cells grown on aromatic compounds or organic acids as the carbon source. Study of [14C]glucose uptake by whole cells indicates that the glucose is taken up by active transport. Metabolic and transport studies clearly demonstrate that glucose metabolism is suppressed when strain CSV86 is grown on aromatic compounds or organic acids.

Preferential utilization of aromatic compounds over glucose by Pseudomonas putida CSV86.

Pseudomonas putida CSV86, a naphthalene-degrading organism, exhibited diauxic growth on aromatic compounds plus glucose, with utilization of aromatics in the first log phase and of glucose in the second log phase. Glucose supplementation did not suppress the activity of degrading enzymes, which were induced upon addition of aromatic compounds. The induction was inhibited by chloramphenicol, suggesting that de novo protein synthesis was essential. Cells showed cometabolism of aromatic compounds and organic acids; however, organic acids suppressed glucose utilization.