Contrastive Open-set Active Learning based Sample Selection for Image Classification.

In this paper, we address a complex but practical scenario in Active Learning (AL) known as open-set AL, where the unlabeled data consists of both in-distribution (ID) and out-of-distribution (OOD) samples. Standard AL methods will fail in this scenario as OOD samples are highly likely to be regarded as uncertain samples, leading to their selection and wasting of the budget. Existing methods focus on selecting the highly likely ID samples, which tend to be easy and less informative. To this end, we introduce two criteria, namely contrastive confidence and historical divergence, which measure the possibility of being ID and the hardness of a sample, respectively. By balancing the two proposed criteria, highly informative ID samples can be selected as much as possible. Furthermore, unlike previous methods that require additional neural networks to detect the OOD samples, we propose a contrastive clustering framework that endows the classifier with the ability to identify the OOD samples and further enhances the network's representation learning. The experimental results demonstrate that the proposed method achieves state-of-the-art performance on several benchmark datasets.

Dynamics of gut microflora across the life cycle of Spodoptera frugiperda and its effects on the feeding and growth of larvae.

BACKGROUND: Spodoptera frugiperda is an important invasive agricultural pest that causes huge economic losses worldwide. Gut microorganisms play a vital role in host feeding, digestion, nutrition, immunity, growth and insecticide resistance. Illumina high-throughput sequencing was used to study the gut microbial community dynamics across the life cycle (egg, 1st to 6th instar larvae, pupae, and male and female adults) of S. frugiperda fed on maize leaves. Furthermore, the gut microbial community and food intake of the 5th instar S. frugiperda larvae were studied after feeding them antibiotics. RESULTS: Enterobacteriaceae and Enterococcaceae dominated the gut during growth and feeding of the larvae. The relative abundance of Enterobacteriaceae was higher in the 4th and 6th instar larvae. With the increase in larval feeding, the relative abundance of Enterococcaceae gradually increased. In addition, principal coordinate analysis and linear discriminant effect size analysis confirmed differences in the structure of gut microbiota at different developmental stages. After antibiotic treatment, the relative abundance of Firmicutes, Proteobacteria and Fusobacteriota decreased. The relative abundance of Enterococcus and Klebsiella decreased significantly. Antibiotic treatment inhibited the gut flora of S. frugiperda, which decreased larval food intake and body weight gain, and prolonged the larval stage. CONCLUSION: The composition of the gut bacterial community plays an important role in the growth, development, and feeding of S. frugiperda. The results have a certain theoretical value for the development of bio-pesticides targeting intestinal flora. © 2022 Society of Chemical Industry.

RNA Sequencing Reveals the Potential Adaptation Mechanism to Different Hosts of Grapholita molesta.

Grapholita molesta is an important fruit tree worldwide pest which feeds on hosts extensively and does serious harm. In this paper, the growth and development parameters and protease activities of G. molesta fed on different hosts were compared. Using Illumina RNA sequencing technology, 18 midgut samples from five different hosts (apple, pear, plum, peach and peach shoots) and artificial diet were sequenced and compared with the reference genome, resulting in 15269 genes and 2785 predicted new genes. From 15 comparative combinations, DEGs were found from 286 to 4187 in each group, with up-regulated genes from 107 to 2395 and down-regulated genes from 83 to 2665. KEGG pathway analysis showed that DEGs were associated with amino acid metabolism, starch and sucrose metabolism, carbohydrate metabolism, and hydrolase activity. A total of 31 co-expression gene modules of different hosts were identified by WGCNA. qRT-PCR showed that the expression pattern of the trypsin gene was consistent with RNA sequencing. In this study, growth and development parameters, protease activity, DEGs, enrichment analysis and qRT-PCR were combined to reveal the adaptation process to different hosts of G. molesta in many aspects. The results of this study provide a basis for further exploration of the molecular mechanism of host adaptation of G. molesta.

Comparison of Gut Bacterial Communities of Fall Armyworm (Spodoptera frugiperda) Reared on Different Host Plants.

Spodoptera frugiperda is a highly polyphagous and invasive agricultural pest that can harm more than 300 plants and cause huge economic losses to crops. Symbiotic bacteria play an important role in the host biology and ecology of herbivores, and have a wide range of effects on host growth and adaptation. In this study, high-throughput sequencing technology was used to investigate the effects of different hosts (corn, wild oat, oilseed rape, pepper, and artificial diet) on gut microbial community structure and diversity. Corn is one of the most favored plants of S. frugiperda. We compared the gut microbiota on corn with and without a seed coating agent. The results showed that Firmicutes and Bacteroidetes dominated the gut microbial community. The microbial abundance on oilseed rape was the highest, the microbial diversity on wild oat was the lowest, and the microbial diversity on corn without a seed coating agent was significantly higher than that with such an agent. PCoA analysis showed that there were significant differences in the gut microbial community among different hosts. PICRUSt analysis showed that most of the functional prediction categories were related to metabolic and cellular processes. The results showed that the gut microbial community of S. frugiperda was affected not only by the host species, but also by different host treatments, which played an important role in host adaptation. It is important to deepen our understanding of the symbiotic relationships between invasive organisms and microorganisms. The study of the adaptability of host insects contributes to the development of more effective and environmentally friendly pest management strategies.

Comparison of Gut Bacterial Communities of Grapholita molesta (Lepidoptera: Tortricidae) Reared on Different Host Plants.

Intestinal symbiotic bacteria have played an important role in the digestion, immunity detoxification, mating, and reproduction of insects during long-term coevolution. The oriental fruit moth, Grapholita molesta, is an important fruit tree pest worldwide. However, the composition of the G. molesta microbial community, especially of the gut microbiome, remains unclear. To explore the differences of gut microbiota of G. molesta when reared on different host plants, we determined the gut bacterial structure when G. molesta was transferred from an artificial diet to different host plants (apples, peaches, nectarines, crisp pears, plums, peach shoots) by amplicon sequencing technology. The results showed that Proteobacteria and Firmicutes are dominant in the gut microbiota of G. molesta. Plum-feeding G. molesta had the highest richness and diversity of gut microbiota, while apple-feeding G. molesta had the lowest. PCoA and PERMANOVA analysis revealed that there were significant differences in the gut microbiota structure of G. molesta on different diets. PICRUSt2 analysis indicated that most of the functional prediction pathways were concentrated in metabolic and cellular processes. Our results confirmed that gut bacterial communities of G. molesta can be influenced by host diets and may play an important role in host adaptation.

Discovery of indolin-2-one derivatives as potent PAK4 inhibitors: Structure-activity relationship analysis, biological evaluation and molecular docking study.

Utilizing a pharmacophore hybridization approach, a novel series of substituted indolin-2-one derivatives were designed, synthesized and evaluated for their in vitro biological activities against p21-activated kinase 4. Compounds 11b, 12d and 12g exhibited the most potent inhibitory activity against PAK4 (IC50=22nM, 16nM and 27nM, respectively). Among them, compound 12g showed the highest antiproliferative activity against A549 cells (IC50=0.83µM). Apoptosis analysis in A549 cells suggested that compound 12g delayed cell cycle progression by arresting cells in the G2/M phase of the cell cycle, retarding cell growth. Further investigation demonstrated that compound 12g strongly inhibited migration and invasion of A549 cells. Western blot analysis indicated that compound 12g potently inhibited the PAK4/LIMK1/cofilin signalling pathways. Finally, the binding mode between compound 12g with PAK4 was proposed by molecular docking. A preliminary ADME profile of the compound 12g was also drawn on the basis of QikProp predictions.

Development of 2, 4-diaminoquinazoline derivatives as potent PAK4 inhibitors by the core refinement strategy.

Upon analysis of the reported crystal structure of PAK4 inhibitor KY04031 (PAK4 IC50 = 0.790 µM) in the active site of PAK4, we investigated the possibility of changing the triazine core of KY04031 to a quinazoline. Using KY04031 as a starting compound, a library of 2, 4-diaminoquinazoline derivatives were designed and synthesized. These compounds were evaluated for PAK4 inhibition, leading to the identification of compound 9d (PAK4 IC50 = 0.033 µM). Compound 9d significantly induced the cell cycle in the G1/S phase and inhibited migration and invasion of A549 cells that over-express PAK4 via regulation of the PAK4-LIMK1 signalling pathway. A docking study of compound 9d was performed to elucidate its possible binding modes and to provide a structural basis for further structure-guided design of PAK4 inhibitors. Compound 9d may serve as a lead compound for anticancer drug discovery and as a valuable research probe for further biological investigation of PAK4.