Performance analysis of large language models in the domain of legal argument mining.

Generative pre-trained transformers (GPT) have recently demonstrated excellent performance in various natural language tasks. The development of ChatGPT and the recently released GPT-4 model has shown competence in solving complex and higher-order reasoning tasks without further training or fine-tuning. However, the applicability and strength of these models in classifying legal texts in the context of argument mining are yet to be realized and have not been tested thoroughly. In this study, we investigate the effectiveness of GPT-like models, specifically GPT-3.5 and GPT-4, for argument mining via prompting. We closely study the model's performance considering diverse prompt formulation and example selection in the prompt via semantic search using state-of-the-art embedding models from OpenAI and sentence transformers. We primarily concentrate on the argument component classification task on the legal corpus from the European Court of Human Rights. To address these models' inherent non-deterministic nature and make our result statistically sound, we conducted 5-fold cross-validation on the test set. Our experiments demonstrate, quite surprisingly, that relatively small domain-specific models outperform GPT 3.5 and GPT-4 in the F1-score for premise and conclusion classes, with 1.9% and 12% improvements, respectively. We hypothesize that the performance drop indirectly reflects the complexity of the structure in the dataset, which we verify through prompt and data analysis. Nevertheless, our results demonstrate a noteworthy variation in the performance of GPT models based on prompt formulation. We observe comparable performance between the two embedding models, with a slight improvement in the local model's ability for prompt selection. This suggests that local models are as semantically rich as the embeddings from the OpenAI model. Our results indicate that the structure of prompts significantly impacts the performance of GPT models and should be considered when designing them.

The Mitogenomes of Ophiostoma minus and Ophiostoma piliferum and Comparisons With Other Members of the Ophiostomatales.

Fungi assigned to the Ophiostomatales are of economic concern as many are blue-stain fungi and some are plant pathogens. The mitogenomes of two blue-stain fungi, Ophiostoma minus and Ophiostoma piliferum, were sequenced and compared with currently available mitogenomes for other members of the Ophiostomatales. Species representing various genera within the Ophiostomatales have been examined for gene content, gene order, phylogenetic relationships, and the distribution of mobile elements. Gene synteny is conserved among the Ophiostomatales but some members were missing the atp9 gene. A genome wide intron landscape has been prepared to demonstrate the distribution of the mobile genetic elements (group I and II introns and homing endonucleases) and to provide insight into the evolutionary dynamics of introns among members of this group of fungi. Examples of complex introns or nested introns composed of two or three intron modules have been observed in some species. The size variation among the mitogenomes (from 23.7 kb to about 150 kb) is mostly due to the presence and absence of introns. Members of the genus Sporothrix sensu stricto appear to have the smallest mitogenomes due to loss of introns. The taxonomy of the Ophiostomatales has recently undergone considerable revisions; however, some lineages remain unresolved. The data showed that genera such as Raffaelea appear to be polyphyletic and the separation of Sporothrix sensu stricto from Ophiostoma is justified.

Comprehensive in silico Study of GLUT10: Prediction of Possible Substrate Binding Sites and Interacting Molecules.

OBJECTIVES: The Arterial Tortuosity Syndrome (ATS) is an autosomal recessive connective tissue disorder, mainly characterized by tortuosity and stenosis of the arteries with a propensity towards aneurysm formation and dissection. It is caused by mutations in the SLC2A10 gene that encodes the facilitative glucose transporter GLUT10. The molecules transported by and interacting with GLUT10 have still not been unambiguously identified. Hence, the study attempts to identify both the substrate binding site of GLUT10 and the molecules interacting with this site. METHODS: As High-resolution X-ray crystallographic structure of GLUT10 was not available, 3D homology model of GLUT10 in open conformation was constructed. Further, molecular docking and bioinformatics investigation were employed. RESULTS AND DISCUSSION: Blind docking of nine reported potential in vitro substrates with this 3D homology model revealed that substrate binding site is possibly made with PRO531, GLU507, GLU437, TRP432, ALA506, LEU519, LEU505, LEU433, GLN525, GLN510, LYS372, LYS373, SER520, SER124, SER533, SER504, SER436 amino acid residues. Virtual screening of all metabolites from the Human Serum Metabolome Database and muscle metabolites from Human Metabolite Database (HMDB) against the GLUT10 revealed possible substrates and interacting molecules for GLUT10, which were found to be involved directly or partially in ATS progression or different arterial disorders. Reported mutation screening revealed that a highly emergent point mutation (c. 1309G>A, p. Glu437Lys) is located in the predicted substrate binding site region. CONCLUSION: Virtual screening expands the possibility to explore more compounds that can interact with GLUT10 and may aid in understanding the mechanisms leading to ATS.

The fungal mitochondrial Nad5 pan-genic intron landscape.

An intron landscape was prepared for the fungal mitochondrial nad5 gene. A hundred and eighty-eight fungal species were examined and a total of 265 introns were noted to be located in 29 intron insertion sites within the examined nad5 genes. Two hundred and sixty-three introns could be classified as group I types and two group II introns were noted. One additional group II intron module was identified nested within a composite group I intron. Based on features related to RNA secondary structures, introns can be classified into different subtypes and it was observed that intron insertion-sites are biased towards phase 0 and they appear to be specific to an intron type. Intron landscapes could be used as a guide map to predict the location of fungal mtDNA mobile introns, which are composite elements that include a ribozyme component and in some instances open reading frames encoding homing endonucleases or reverse transcriptases and all of these have applications in biotechnology.

The mitochondrial genome of Endoconidiophora resinifera is intron rich.

Endoconidiophora resinifera (=Ceratocystis resinifera) is a blue-stain fungus that occurs on conifers. The data showed that the Endoconidiophora resinifera mitochondrial genome is one of the largest mitochondrial genomes (>220 kb) so far reported among members of the Ascomycota. An exceptional large number of introns (81) were noted and differences among the four strains were restricted to minor variations in intron numbers and a few indels and single nucleotide polymorphisms. The major differences among the four strains examined are due to size polymorphisms generated by the absence or presence of mitochondrial introns. Also, these mitochondrial genomes encode the largest cytochrome oxidase subunit 1 gene (47.5 kb) reported so far among the fungi. The large size for this gene again can be attributed to the large number of intron insertions. This study reports the first mitochondrial genome for the genus Endoconidiophora, previously members of this genus were assigned to Ceratocystis. The latter genus has recently undergone extensive taxonomic revisions and the mitochondrial genome might provide loci that could be applied as molecular markers assisting in the identification of taxa within this group of economically important fungi. The large mitochondrial genome also may provide some insight on mechanisms that can lead to mitochondrial genome expansion.

The complete mitochondrial genome of the Dutch elm disease fungus Ophiostoma novo-ulmi subsp. novo-ulmi.

Ophiostoma novo-ulmi, a member of the Ophiostomatales (Ascomycota), is the causal agent of the current Dutch elm disease pandemic in Europe and North America. The complete mitochondrial genome (mtDNA) of Ophiostoma novo-ulmi subsp. novo-ulmi, the European component of O. novo-ulmi, has been sequenced and annotated. Gene order (synteny) among the currently available members of the Ophiostomatales was examined and appears to be conserved, and mtDNA size variability among the Ophiostomatales is due in part to the presence of introns and their encoded open reading frames. Phylogenetic analysis of concatenated mitochondrial protein-coding genes yielded phylogenetic estimates for various members of the Ophiostomatales, with strong statistical support showing that mtDNA analysis may provide valuable insights into the evolution of the Ophiostomatales.

Functional characterization of an N-terminally-truncated mitochondrial porin expressed in Neurospora crassa.

Mitochondrial porin, which forms voltage-dependent anion-selective channels (VDAC) in the outer membrane, can be folded into a 19-ß-stranded barrel. The N terminus of the protein is external to the barrel and contains α-helical structure. Targeted modifications of the N-terminal region have been assessed in artificial membranes, leading to different models for gating in vitro. However, the in vivo requirements for gating and the N-terminal segment of porin are less well-understood. Using Neurospora crassa porin as a model, the effects of a partial deletion of the N-terminal segment were investigated. The protein, ΔN2-12porin, is assembled into the outer membrane, albeit at lower levels than the wild-type protein. The resulting strain displays electron transport chain deficiencies, concomitant expression of alternative oxidase, and decreased growth rates. Nonetheless, its mitochondrial genome does not contain any significant mutations. Most of the genes that are expressed in high levels in porin-less N. crassa are expressed at levels similar to that of wild type or are slightly increased in ΔN2-12porin strains. Thus, although the N-terminal segment of VDAC is required for complete function in vivo, low levels of a protein lacking part of the N terminus are able to rescue some of the defects associated with the absence of porin.

In vitro screening for inhibitor of cloned Drosophila melanogaster tyramine-ß-hydroxylase and docking studies.

Biogenic amines are common biologically active substances extended within the whole animal kingdom where they play vital roles as signal transducer as well as regulator of cell functions. One of these biogenic amines called octopamine (OA) is synthesized from tyramine (TA) by the catalysis of tyramine-ß-hydroxylase (TßH) originated in the insect nervous system. Both TA and OA act as neurotransmitters, neurohormones and neuromodulators in the arthropod nervous system. Herein, the inhibitory activity of 1-arylimidazole-2(3H)-thiones (AITs) was tested on cloned Drosophila tyramine-ß-hydroxylase (DmTßH) expressed in Bombyx mori strain. Radiolabelled 3H-TA was used to analyze the activity of AITs exhibited inhibitory effects on DmTßH, whose ID50 values range from 0.02 to 2511nM where DmTßH was inhibited in a dose-dependent manner at pH 7.6 and 25°C during a 30min of incubation. To understand the catalytic role of the TßH, a three dimensional structure of the TßH from Drosophila melanogaster was constructed by homology modeling using the Phyre2 web server with 100% confidence. The modeled three-dimensional structure of TßH was used to perform the docking study with AITs. This may give more insights to precise design of inhibitors for TßH to control insect's population.

Molecular docking simulations provide insights in the substrate binding sites and possible substrates of the ABCC6 transporter.

The human ATP-binding cassette family C member 6 (ABCC6) gene encodes an ABC transporter protein (ABCC6), primarily expressed in liver and kidney. Mutations in the ABCC6 gene cause pseudoxanthoma elasticum (PXE), an autosomal recessive connective tissue disease characterized by ectopic mineralization of the elastic fibers. The pathophysiology underlying PXE is incompletely understood, which can at least partly be explained by the undetermined nature of the ABCC6 substrates as well as the unknown substrate recognition and binding sites. Several compounds, including anionic glutathione conjugates (N-ethylmaleimide; NEM-GS) and leukotriene C4 (LTC4) were shown to be modestly transported in vitro; conversely, vitamin K3 (VK3) was demonstrated not to be transported by ABCC6. To predict the possible substrate binding pockets of the ABCC6 transporter, we generated a 3D homology model of ABCC6 in both open and closed conformation, qualified for molecular docking and virtual screening approaches. By docking 10 reported in vitro substrates in our ABCC6 3D homology models, we were able to predict the substrate binding residues of ABCC6. Further, virtual screening of 4651 metabolites from the Human Serum Metabolome Database against our open conformation model disclosed possible substrates for ABCC6, which are mostly lipid and biliary secretion compounds, some of which are found to be involved in mineralization. Docking of these possible substrates in the closed conformation model also showed high affinity. Virtual screening expands this possibility to explore more compounds that can interact with ABCC6, and may aid in understanding the mechanisms leading to PXE.

Palindromes drive the re-assortment in Influenza A.

Different subtypes of Influenza A virus are associated with species specific, zoonotic or pandemic Influenza. The cause of its severity underlies in complicated evolution of its segmented RNA genome. Although genetic shift and genetic drift are well known in the evolution of this virus, we reported the significant role of unique RNA palindromes in its evolution. Our computational approach identified the existence of unique palindromes in each subtype of Influenza A virus with its absence in Influenza B relating the fact of virulence and vigorous genetic hitchhiking in Influenza A. The current study focused on the re-assortment event responsible for the emergence of pandemic-2009 H1N1 virus, which is associated with outgrow of new palindrome and in turn, changing its RNA structure. We hypothesize that the change in RNA structure due to the presence of palindrome facilitates the event of re-assortment in Influenza A. Thus the evolutionary process of Influenza A is much more complicated as previously known, and that has been demonstrated in this study.

Envelope Proteins Pertain with Evolution and Adaptive Mechanism of the Novel Influenza A/H1N1 in Humans.

The novel swine-origin influenza A/H1N1 virus (S-OIV) first detected in April 2009 has been identified to transmit from human to human directly and is the cause of currently emerged pandemic. In this study, nucleotide and deduced amino acid sequences of hemagglutinin (HA) and neuraminidase (NA) of the S-OIV and other influenza A viruses were analyzed through bioinformatic tools for phylogenetic analysis, genetic recombination and point mutation to investigate the emergence and adaptation of the S-OIV in human. The phylogenetic analysis showed that the HA comes from triple reassortant influenza A/H1N2 and the NA from Eurasian swine influenza A/H1N1 indicating HA and NA to descend from different lineages during the genesis of the S-OIV. Recombination analysis nullified the possibility of occurrence of recombination in HA and NA denoting the role of reassortment in the outbreak. Several conservative mutations are observed in the amino acid sequences of the HA and NA and this mutated residues are identical in the S-OIV. The results reported herein suggested the notion that the recent pandemic is the result of reassortment of different genes from different lineages of two envelope proteins, HA and NA which are responsible for antigenic activity of virus. This study further suggests that the adaptive capability of the S-OIV in human is acquired by the unique mutations generated during emergence.