Background: Plants and their associated microbiota constitute an assemblage of species known as holobionts. The plant seed microbiome plays an important role in nutrient uptake and stress attenuation. However, the core vertically transmitted endophytes remain largely unexplored. Results: To gain valuable insights into the vertical transmission of rice seed core endophytes, we conducted a large-scale analysis of the microbiomes of two generations of six different rice varieties from five microhabitats (bulk soil, rhizosphere, root, stem, and seed) from four geographic locations. We showed that the microhabitat rather than the geographic location and rice variety was the primary driver of the rice microbiome assemblage. The diversity and network complexity of the rice-associated microbiome decreased steadily from far to near the roots, rice exterior to interior, and from belowground to aboveground niches. Remarkably, the microbiomes of the roots, stems, and seeds of the rice interior compartments were not greatly influenced by the external environment. The core bacterial endophytes of rice were primarily comprised of 14 amplicon sequence variants (ASVs), 10 of which, especially ASV_2 (Pantoea) and ASV_48 (Xanthomonas), were identified as potentially vertically transmitted taxa because they existed across generations, were rarely present in exterior rice microhabitats, and were frequently isolated from rice seeds. The genome sequences of Pantoea and Xanthomonas isolated from the parental and offspring seeds showed a high degree of average nucleotide and core protein identity, indicating vertical transmission of seed endophytes across generations. In silico prediction indicated that the seed endophytes Pantoea and Xanthomonas possessed streamlined genomes with short lengths, low-complexity metabolism, and various plant growth-promoting traits. We also found that all strains of Pantoea and Xanthomonas exhibited cellulase activity and produced indole-3-acetic acid. However, most strains exhibited insignificant antagonism to the major pathogens of rice, such as Magnaporthe oryzae and X. oryzae pv. oryzae. Conclusion: Overall, our study revealed that microhabitats, rather than site-specific environmental factors or host varieties, shape the rice microbiome. We discovered the vertically transmitted profiles and keystone taxa of the rice microbiome, which led to the isolation of culturable seed endophytes and investigation of their potential roles in plant-microbiome interactions. Our results provide insights on vertically transmitted microbiota and suggest new avenues for improving plant fitness via the manipulation of seed-associated microbiomes.

Plant growth-promoting rhizobacterial strain FP607^T was isolated from the rhizosphere of beets in Wuhan, China. Strain FP607^T exhibited significant antagonism toward several phytopathogenic bacteria, indicating that FP607^T may produce antimicrobial metabolites and has a stronger biocontrol efficacy against plant pathogens. Growth-promoting tests showed that FP607^T produced indole-3-acetic acid (IAA), NH₃, and ferritin. The genome sequence of strain FP607^T was 6,590,972 bp long with 59.0% G + C content. The optimum temperature range was 25–30 °C, and the optimum pH was 7. The cells of strain FP607^T were Gram-negative, short, and rod-shaped, with polar flagella. The colonies on the King’s B (KB) agar plates were light yellow, smooth, and circular, with regular edges. A phylogenetic analysis of the 16S rRNA sequence and a multilocus sequence analysis (MLSA) showed that strain FP607^T was most closely related to the type of strain Pseudomonas farris SWRI79^T. Based on a polyphasic taxonomic approach, strain FP607^T was identified as a novel species within the genus Pseudomonas, for which the name Pseudomonas wuhanensis sp. nov. was proposed. The type of strain used was FP607^T (JCM 35688, CGMCC 27743, and ACCC 62446).

Three bacterial strains, FP250^T, FP821, and FP53, were isolated from the rhizosphere soil of oilseed rape, licorice, and habanero pepper in Anhui Province, Xinjiang Uygur Autonomous Region, and Jiangsu Province, PR China, respectively. All strains were shown to grow at 4–37 °C and pH 6.0–9.0, and in the presence of 0–4.0 % (w/v) NaCl. Phylogenetic analyses based on 16S rRNA gene sequences or housekeeping genes (16S rRNA, gyrB, rpoB, and rpoD) and phylogenomic analysis showed that strains FP250^T, FP821, and FP53 belong to the genus Pseudomonas, and are closely related to Pseudomonas kilonensis DSM 13647^T, Pseudomonas brassicacearum JCM 11938^T, Pseudomonas viciae 11K1^T, and Pseudomonas thivervalensis DSM 13194^T. The DNA G+C content of strain FP205^T was 59.8 mol%. The average nucleotide identity and digital DNA–DNA hybridization values of strain FP205^T with the most closely related strain were 93.2 % and 51.4 %, respectively, which is well below the threshold for species differentiation. Strain FP205^T contained summed feature 3 (C_16 : 1  ω6c and/or C_16 : 1  ω7c), summed feature 8 (C_18 : 1  ω7c and/or C_18 : 1  ω6c) as major fatty acids, and diphosphatidylglycerol along with phosphatidylethanolamine and aminophospholipid as major polar lipids. The predominant isoprenoid quinone was ubiquinone-9. Based on these phenotypic, phylogenetic, and chemotaxonomic results, strain FP205^T represents a novel species of the genus Pseudomonas, for which the name Pseudomonas hefeiensis sp. nov. is proposed. The type strain is FP205^T (=ACCC 62447^T=JCM 35687^T).

Knowledge classification is a critical problem that has to be solved in implementation of knowledge management. However, a highlight issue is that the existing knowledge classification methods are unpractical for enterprise. In order to realize the automatic classification of enterprise knowledge, a method of enterprise text knowledge classification based on knowledge schema is proposed, which describes the enterprise text knowledge and management objects with knowledge schema, and matches knowledge schémas extracted from texts with meta-knowledge, and then matches the formed enterprise knowledge schémas with knowledge schémas of management objects and sorts the matching results. A knowledge classification system based on this method was developed. The experiment results show that this method is more practical and its classification accuracy can satisfy the requirements in practical applications.