Phylotaxogenomics for the Reappraisal of the Genus Roseomonas With the Creation of Six New Genera

The genus Roseomonas is a significant group of bacteria which is invariably of great clinical and ecological importance. Previous studies have shown that the genus Roseomonas is polyphyletic in nature. Our present study focused on generating a lucid understanding of the phylogenetic framework for the re-evaluation and reclassification of the genus Roseomonas. Phylogenetic studies based on the 16S rRNA gene and 92 concatenated genes suggested that the genus is heterogeneous, forming seven major groups. Existing Roseomonas species were subjected to an array of genomic, phenotypic, and chemotaxonomic analyses in order to resolve the heterogeneity. Genomic similarity indices (dDDH and ANI) indicated that the members were well-defined at the species level. The Percentage of Conserved Proteins (POCP) and the average Amino Acid Identity (AAI) values between the groups of the genus Roseomonas and other interspersing members of the family Acetobacteraceae were below 65 and 70%, respectively. The pan-genome evaluation depicted that the pan-genome was an open type and the members shared 958 core genes. This claim of reclassification was equally supported by the phenotypic and chemotaxonomic differences between the groups. Thus, in this study, we propose to re-evaluate and reclassify the genus Roseomonas and propose six novel genera as Pararoseomonas gen. nov., Falsiroseomonas gen. nov., Paeniroseomonas gen. nov., Plastoroseomonas gen. nov., Neoroseomonas gen. nov., and Pseudoroseomonas gen. nov.

Phylotaxogenomics of the class Alphaproteobacteria was extensively studied recently by Hördt et al. (2020), which showed that several taxa under it, including the members of the family Acetobacteraceae, were disordered. One of the major discussions under this communication is the re-evaluation of the phylogenetic relationships between the members of the genus Roseomonas. This is based on the fact that the genus Roseomonas was distinctly polyphyletic with members like Roseomonas stagni and Roseomonas lacus interspersed in between other sister groups of Humitalea and Rubritepida in the family Acetobacteraceae (Hördt et al., 2020). Furthermore, the same study revealed that Rhodovarius lipocyclicus nested within the genus Roseomonas based on the constrained comprehensive tree (CCT) (Hördt et al., 2020). Likewise, a study conducted by Romano-Bertrand et al. (2016) depicted that the members of the genus Roseomonas were organized into seven different clades based on the 16S rRNA gene phylogeny and ecology.
Ambiguity has emerged due to the poor discriminatory power of the16S rRNA gene marker for absolute resolution at the genus/species level (Fox et al., 1992;Janda and Abbott, 2007;Olm et al., 2020). Over the decade, meaningful deductions of unresolved and longstanding phylogenetic relationships between members of the genera like Mycobacterium (Gupta et al., 2018), Corallococcus (Livingstone et al., 2018), Roseobacter (Wirth and Whitman, 2018), Rhodobacter (Suresh et al., 2019), and 1 https://lpsn.dsmz.de/family/acetobacteraceae 2 https://lpsn.dsmz.de/genus/roseomonas Lactobacillus (Wittouck et al., 2019) have been worked out well. This resolution has been successfully elucidated owing to the robust calculated parameters and integrated comparative genomics that included average amino acid identity (AAI), average nucleotide identity (ANI), genomic signatures, and pangenome analysis (McInerney et al., 2017). In this study, we revisited the current status of the genus Roseomonas based on definite and congruent genomic evidence. Here, on the basis of phylogenomic studies like phylogenetic, pan-genomic, and taxogenomic analysis, we have proposed to re-classify the genus Roseomonas into Roseomonas genus sensu-stricto along with six novel genera (gen. nov.).

Genome Sequences and Phylogenetic Analysis
The genome sequences affiliated to the family Acetobacteraceae, which were available publicly at the NCBI and JGI databases, were retrieved (Supplementary Table 1). A total of 34 type strains and 28 other related strains affiliated to the genus Roseomonas were considered in this study ( Table 1). As the genome sequence of the type species of the genus Roseomonas was not available at the time of writing, R. gilardii subsp. rosea was taken as a representative for the type species in this analysis.
For the phylogenetic analysis, 16S rRNA gene sequences of type strains of all the validly published species names, effectively but not validly published names like "R. aceris" R-1 T , "R. chloroacetimidivorans" BUT-13 T , and "candidatus Roseomonas massiliae" of the genus Roseomonas, were taken from the NCBI database. For Roseomonas aquatica and Roseomomas fluminis, cloned sequences were also taken to understand the phylogenetic position and confirm their status within the genus Roseomonas. In addition, the 16S rRNA gene sequences of representative members and cloned sequences of 11 other genera affiliated to the family of Acetobacteraceae were also accounted for the phylogenetic analysis. The 16S rRNA sequence of Elioraea tepidiphila DSM 17972 T was used as an outgroup. MUSCLE algorithm (Edgar, 2004) of MEGA7 was used for sequence alignments and the phylogenetic analysis of the sequences was performed using MEGA7 (Kumar et al., 2016). Distances were calculated using Kimura two-parameter in a deletion manner (Kimura, 1980). Neighbor-joining (NJ), maximum likelihood (ML), and maximum parsimony (MP) methods in the MEGA7 software were used to reconstruct phylogenetic trees. Percentage support values were obtained using a bootstrap procedure (Felsenstein, 1985).
As for the phylogenomic tree reconstruction, 62 genomes of members of the genus Roseomonas (34 type strains, 28 other related strains) and 135 genomes representative of 43 other genera of the family Acetobacteraceae (as per the NCBI database) were considered for the phylogenomic study (Supplementary Table 1). A total of 92 core genes were identified in 197 genomes and were retrieved for phylogenomic analysis using the Up-todate Bacterial Core Gene (UBCG) tool (Na et al., 2018). The genes considered for the phylogenomic tree analysis are given  in Supplementary Table 2, and Stella humosa DSM 5900 T was used as an outgroup. The RAXML-based phylogenomic tree was constructed using the concatenated sequences of the 92 core genes of 197 genomes.

Analysis of Core and Pan-Genome
Bacterial Pan-genome Analysis (BPGA) pipeline (Chaudhari et al., 2016) was applied for the analysis of the genomic diversity of Roseomonas. The default parameters were set to check the conserved, accessory, and strain-specific genomic traits between the members of the genus Roseomonas. The same process was also carried out for each of the proposed genera (n ≥ 3) in this study.

Genomic Similarity Indices
Robust parameters like average nucleotide identity (ANI) and digital DNA-DNA hybridization (dDDH) were computed for the precise delineation of the members of the genus Roseomonas. Whereas, AAI and Percentage of Conserved Proteins (POCP) values were calculated within the members of the genus Roseomonas as well as with the members of the other genera in the family Acetobactereaceae. OrthoANI tool (Yoon et al., 2017) and Genome-to-Genome Distance calculator 2.1 3 (Auch et al., 2010) were used for calculating the ANI and dDDH values, respectively. AAI was calculated using the AAI calculator developed by Kostas lab 4 . POCP was calculated as described by Qin et al. (2014). The obtained AAI and 3 http://ggdc.dsmz.de/ggdc.php# 4 http://enve-omics.ce.gatech.edu/aai/ POCP values were used to construct a heatmap using an online tool, Morpheus, 5 and were clustered hierarchically based on the Euclidean distance.

Functional and Metabolic Annotations
The identification and annotation of biosynthetic gene clusters (BCGs) for the members of the genus Roseomonas were carried out with the antiSMASH 5.0 web server (Blin et al., 2019) which applies the Hidden Markov Models for the identification of BCGs. For the detection, strictness level was set to "relaxed." Genomes were submitted to the KBase online software (Arkin et al., 2018) 6 using the default parameters to generate genomic data which were later manually analyzed to gather the groupspecific gene clusters.

Virulence Factors and Pathogen-Associated Genes
To understand the pathogenicity of the members of the genus Roseomonas, genomes were submitted individually to the IslandViewer 4 server (Bertelli et al., 2017) in the .gbk format where Salmonella enterica Serovar Typhimurium LT2 was given as the reference genome. Virulence factors were also checked by submitting the genome sequences (.fas format) to the VirulenceFinder 2.0 7 hosted by the Center for Genomic Epidemiology (CGE) against four genera of Listeria, Staphylococcus, Escherichia, and Enterococcus. The threshold for percent identity (%ID) between input and matching gene in the database was 90% with a minimum length of 60%.

Phenotypic and Chemotaxonomic Characters
Phenotypic and chemotaxonomic characterization of each described species for taxon delineation was obtained from the original species descriptions, Bergey's Manual of Systematic Bacteriology (Weyant and Whitney, 2005), and from other references as specified.

RESULTS
Phylogenetic Analysis Based on the 16S rRNA Gene and Genome Sequence The 16S rRNA gene-based phylogenetic tree showed that the genus Roseomonas was polyphyletic and was segregated into seven major clades (Figure 1). These seven major clades were defined as Groups I-VII on the basis of the clade formation.
The validly described members of each of the clade were as follows: Group I (n = 4), Group II (n = 8), Group III (n = 10), Group IV (n = 2), Group V (n = 2), Group VI (n = 7), and Group VII (n = 16) (Figure 1). Group I shall be considered as Roseomonas genus sensu stricto. The pairwise 16S rRNA gene identity was calculated using the pairwise nucleotide sequence alignment for taxonomy. 8 The members of a group delineated by the phylogenetic tree shared at least 93.5-99.2% of the 16S rRNA gene identity (Supplementary Table 3).
The phylogenomic tree was reconstructed using 92 core genes (Supplementary Table 2) of 197 genomes by the UBCG tool (Na et al., 2018). The study by Na et al. (2018) showed that these 92 core genes (Supplementary Table 2) were found as a single copy in more than 95% of the genome sequences. Furthermore, the UBCG-based phylogenomic tree (92 genes) could separate E. coli/Shigella spp. better than the 16S rRNA gene alone (Na et al., 2018). UBCG-based phylogenomic tree (Figure 2) further confirmed the heterogeneity of the genus Roseomonas. Genome sequences were not available for the members of Group IV including R. aquatica and R. fluminis. Furthermore, as per this study, Roseomonas aeriglobus and a genomospecies of Roseomonas cladded outside of the family Acetobacteraceae. Therefore, the genomes of these members were not considered for further study.

Analysis of Core and Pan-Genome of the Genus Roseomonas
To examine the distribution of genes and genomic diversity across the genus Roseomonas, genome sequences were given as input in the BPGA tool. The analyzed data of the genus Roseomonas are given in Supplementary Table 4. Genus Roseomonas members have 958 core genes (19.1%), 219,753 accessory genes (72.8%), and 24,320 strain-specific genes (8.1%) ( Figure 3A). For the diagrammatic representation of the pangenome for the genus Roseomonas, only the type strains were considered. The core-pan plot ( Figure 3B) showed an open pangenome for the genus Roseomonas as it did not level off into a plateau and extended with the increase in the number of genomes. The core genome was conserved at the genus level as the plot leveled off. The KEGG distribution (%) of genes based on function is illustrated in Figure 3C for the genus Roseomonas. Members of Groups I, II, III, V, VI, and VII had core genes between the range of 1,573-3,152 (Supplementary Table 4 Genomic Metrics-dDDH, ANI, AAI, and POCP At an intra-group level, members of a group (consisting both type and non-type strains) defined by phylotaxogenomics shared dDDH and ANI values of at least 21-100% and 78.0-99.9%, respectively (Supplementary Table 5). At an inter-group level, members of different groups did not show dDDH and ANI values above 20.8 and 76.8%, respectively. POCP and AAI values were calculated between the members of the groups (intra-group) as well as at the inter-group level (Supplementary Tables 6, 7) as they are considered pivotal and accurate for genus delineation (Qin et al., 2014). POCP values for Group I (n = 11), Group II (n = 9), Group III (n = 8), Group V (n = 3), Group VI (n = 12), and Group VII (n = 17) taxa were 80-98.5%, 65.  (Supplementary Figures 3, 4) support the division of the six groups. The AAI values between Roseomonas fauriae (later heterotypic synonym of Azospirillum brasilense) and the members of Group I are in between 51.2 and 51.5%, thus making them distinct species under different genera.
As for the Group VII, four genomes had unique ribosomally synthesized and post-translationally modified peptides (RiPPs) BCG exclusively.
The genome annotation showed that 226, 58, and 82 gene clusters were unique to Groups I, II, and III, respectively (Supplementary Table 9). Members of Group V comprised of 67 unique gene clusters, Group VI consisted of 74 unique gene clusters, and Group VII comprised16 unique gene clusters (Supplementary Table 9). As predicted by the IslandViewer 4 server and VirulenceFinder 2.0, the draft genomes of the members of the genus Roseomonas did not possess any genes related to its pathogenicity and virulence.

Phenotypic and Chemotaxonomic Characters Between the Groups
Correlation between the genomic studies and phenotypic characters are desirable for supporting taxa delineation. Members of Groups II and VI can be distinguished from those of Group V by being non-motile. Members of Groups II and III can be differentiated from those of Groups I, IV, V, VI, and VII in showing variable catalase activity. Members of Groups I and II show variable oxidase activity, whereas members of FIGURE 2 | Phylogenomic tree constructed using 92 core genes tool based on the Up-to-date Bacterial Core Gene (UBCG) (Na et al., 2018). The tree was generated using the MEGA7 software (NJ) with Stella humosa DSM 5900 T (RJKX00000000) as an outgroup.
the other groups are either oxidase positive (Groups III, VI, VII) or negative (Groups IV, V). Members of Groups VI and VII have a high NaCl tolerance than members of the other groups. NaCl tolerance was used as a differentiating taxonomic character for delineating the genus Swaminathania biochemically from the other genera of the family Acetobacteraceae like  Analysis of polar lipids is a significant chemotaxonomic aid and has often been used as a differential character for reclassification.
The phenotypic differentiating characteristics between the genus Roseomonas and other closely related genera of the family Acetobacteraceae are given in Table 2.
Frontiers in Microbiology | www.frontiersin.org standing between the current members of genus Roseomonas, a comprehensive taxogenomic analysis was carried out. The 16S rRNA gene-based phylogenetic tree showed that the genus Roseomonas was polyphyletic and was segregated into seven major clades: Groups I to VII (Figure 1). The cladding patterns of Groups I, II, III, VI, and VII members in this study are comparable to the cladding pattern observed by Romano-Bertrand et al. (2016) in terms of the species composition. Group IV member, R. aquatica (Group IV), cladded outside the designated clade observed by Romano-Bertrand et al. (2016) whereas, R. fluminis which also belongs to this group was described later in 2018 by Ko et al. (2018). Furthermore, R. arctica (Group V) was shown to be affiliated to the R. stagni (Group III) clade, whereas "R. hellenica" is not included by them as it was described later (Rat et al., 2021). It was suggested that the whole genome phylogenies could provide a better resolution rather than a single gene like the 16S rRNA gene-based phylogenies for taxonomic delineation (Logan et al., 2009). However, in the case of Roseomonas, the phylogenomic tree based on 92 concatenated genes also showed the same cladding pattern (Figure 2; except for Group IV for which the genome sequences are not available). Nouioui et al. (2018) and Hördt et al. (2020) have shown that genome size variation can be applied as a reliable taxonomic marker. They showed that the genome size appeared to be genus-specific for the members of the phylum Actinobacteria and the class Alphaproteobacteria. Both studies implied that the genome size and G+C content (mol%) were phylogenetically conserved. Hence, the genome sizes have been formally added to the descriptions. In this context, Group I had the smallest genome size as compared to the other groups. Analysis of pan and core genomes using the BPGA pipeline illustrated that the pan-genome of the genus Roseomonas is open and consists of 958 core genes, summing up to 19.1% of the total pan-genome (Supplementary Table 3). As for the BCGs, only groups of the genus Roseomonas could produce specific secondary metabolites, i.e., NRPS, fengymycin, and asukamycic (Group I); redoxcofactor (Group II); and NRPS/T1PKS and arylpolene (Group V) (Supplementary Table 8). The genome annotation also revealed the composition of the gene clusters unique to each group with the highest number of clusters unique to Group I and lowest in Group VII (Supplementary Table 9). Chun et al. (2018) have proposed minimum standards based on an overall genome-related index (OGRI) like dDDH and ANI for species delineation. Wayne et al. (1984) prescribed a cutoff of 70% dDDH value for species delineation. In the present scenario, ≥95% of ANI between the two strains concludes that both belong to the same species whereas for values <95%, two strains are considered as different species (<75% for different genera) (Richter and Rosselló-Móra, 2009;Rodriguez and Konstantinidis, 2014;Rosselló-Móra and Amann, 2015). Both indices (dDDH and ANI) are consistent at the intragroup level with the recommended standards confirming that all members of Roseomonas are well-described at the species level (Supplementary Table 3). For the POCP values, <50% were considered a cut-off for genera delineation (Qin et al., 2014). POCP values between the groups of the genus Roseomonas were <65% (40-65%) (Figure 4 and Supplementary Table 6) and with the other genera of the members of the family Acetobacteraceae were also <70% (20-60%) ( Supplementary  Table 6). Thus, for the species of the genus Roseomonas, the calculated values of POCP for genus delineation were not in agreement with the ones observed by Qin et al. (2014). Although a proposed genus boundary of the POCP value for prokaryotic lineages was assigned as 50%, many studies later showed exceptions at the inter-genera comparison. The universal cut-off of 50% is considered conservative, as being only an index of relatedness  as reflected in its ineffectiveness in delineating different genera of the family Methylococcaceae (Orata et al., 2018), Bacillaceae (Aliyu et al., 2016), Burkholderiaceae (Lopes-Santos et al., 2017), and Rhodobacteraceae (Wirth and Whitman, 2018). Luo et al. (2014) have shown that AAI values of related but different genera ranged from 60 to 80%. In our study, AAI values were below 70% between the different groups of the genus Roseomonas (Figure 4 and Supplementary Table 7) and other genera members of the family Acetobacteraceae (Supplementary Table 7), hence in congruence with the work of Luo et al. (2014).
The study conducted by Hördt et al. (2020) showed that certain members of the genus Roseomonas like R. stagni were interspersed with H. rosea and R. lacus with Rubritepida flocculans. The distinction between the mentioned species pairs was clarified by the dDDH, ANI, AAI, and POCP values. dDDH, ANI, AAI, and POCP values between R. stagni and H. rosea and R. lacus and R. flocculans were 19.4, 74.1, 64.2, and 64.2 and 19.8, 74.8, 66.6, and 57.1%, respectively. Similarly, the study also showed that Rhodovarius lipocyclicus was nested within the genus Roseomonas (Hördt et al., 2020). However, R. lipocyclicus is a distinct member of the genus Rhodovarius from Roseomonas as the AAI and POCP values are 50.8-66.8% and 53.2-64.7%, respectively, below the recommended cut-off for genus delineation. Thus, the above indices clearly differentiate the Roseomonas species from the genera of Humitalea, Rubritepida, and Rhodovarius. Therefore, the above detailed discussion entails the formation of seven different groups within the genus Roseomonas in congruence with the findings given by Romano-Bertrand et al. (2016) and Hördt et al. (2020).
Members of the genus Roseomonas have a ubiquitous distribution in the environment. Group I members were the only ones to have been isolated from a human blood sample (R. gilardii, R. mucosa, R. gilardii subsp. rosea) (Rihs et al., 1993;Han et al., 2003) except for R. elaeocarpi which was isolated from olive phyllosphere (Damtab et al., 2016). An examination of the literature demonstrated that the strains of Roseomonas have been widely reported in human infections and isolated from clinical samples like sputum, wounds, and genitourinary sites (Wallace et al., 1990;Rihs et al., 1993). It was also found to be associated with immunocompromised patients (Marin et al., 2001). Other reports of bacteremia caused by the Roseomonas species were in vertebral osteomyelitis (Nahass et al., 1995), infected tooth (Diesendorf et al., 2017), endocarditis (Shao et al., 2019), and peritonitis (Malini et al., 2016). However, many reports in case studies suggests that the majority of the patients had underlying diseases, malignancy being the most common factor (Wang et al., 2012;Michon et al., 2014). Pathogenic genes and virulence factors are potent segments of the genomic islands. They become part of the genome as a consequence of horizontal gene transfer, and these genes are observed to confer their pathogenicity and virulence to the bacteria (da Silva Filho et al., 2018). The abovementioned reports can be substantiated with the predicted facts that the genomes of the members do not suggest any kind of genes responsible toward its pathogenicity and virulence. Thus, it may be concluded that Roseomonas strains are incidental and not causative of pathogenicity. In addition to being associated with human clinical samples, Roseomonas spp. were also isolated from various sources like soil (Kim and Ka, 2014), sediment , sludges , contaminated soil (Subhash and Lee, 2018), freshwater (Baik et al., 2012), and also human sources like blood (Rihs et al., 1993; Figure 1).
In an ecological context, the presence of accessory genes in the members shows their divergence due to environmental adaptations. The presence of accessory genes shows the acquisition of genes in response to the selective pressure (Brito et al., 2015) or for the colonization of the new ecological habitats (McInerney et al., 2017). The cosmopolitan distribution of the members of the groups may be attributed to the gain of accessory genes for their survival which in turn may be attributed to their variable genomic sizes, simple organization, or horizontal gene transfer rates (Kuo and Ochman, 2009;Aherfi et al., 2018). Also, the formation of ecotypes based on the varied isolation sources may not be feasible for groups of the genus Roseomonas. This is also evident from the distinctive genomic features observed during the pan-genome evaluation for each of the groups, which may have resulted in response to the environmental changes.
The extensive taxogenomics study performed substantiates the separation of the respective groups of the genus Roseomonas into separate genera as suggested by Romano-Bertrand et al. (2016) and Hördt et al. (2020). Considering the fact that Roseomonas is a very divergent genus, henceforth members of this genus cannot be clustered under one phylogenetic genus, necessitating the creation of six new genera. For each of the novel genera, emended description and reclassification are given. Phylogenetic trees based on the 16S rRNA gene (Figure 1) and whole genome sequences (Figure 2) agree with the formation of seven different clades as separate genera. Although the 16S rRNA gene studies have certain drawbacks, this approach is still quite accountable for the delineation of the members of the genus Roseomonas. This study would be a first attempt in re-evaluating the genus Roseomonas and reclassifying it by creating six new genera: Group II as Pararoseomonas gen. nov., Group III as Falsiroseomonas gen. nov., Group IV as Paeniroseomonas gen. nov., Group V as Plastoroseomonas gen. nov., Group VI as Neoroseomonas gen. nov., and Group VII as Pseudoroseomonas gen. nov. The genus delineation for Group IV is based on the 16S rRNA gene analysis (in the absence of genome sequences) and phenotypic characteristics. The genus delineation constructed from the 16S rRNA gene sequences and phenotypic characters provide sufficient resolution to distinguish the Group IV members from other groups of the genus Roseomonas. However, the availability of the genome sequences may further substantiate the reclassification in the future. The species within each of the newly defined genera have been defined accurately as per the recommended standards for ANI and dDDH (Supplementary Table 5). As for the genus delineation, the inter-group AAI and POCP values agree with the standards given for genus determination (Supplementary Tables 6, 7). The justification for genus delineation based on the 16S rRNA gene identity, AAI and POCP, and other phenotypic characters are given in Figure 4 and Table 2.
Emended Description of the Genus Roseomonas by Rihs et al. (1998) The description is the same as the one given by Rihs et al. (1993) and Sánchez-Porro et al. (2009) for the genus Roseomonas except for a few modifications. H 2 S formation is variable and the genome-based G+C content is ∼70 mol%.
The type species is Roseomonas gilardii.
Description of Pararoseomonas gen. nov.
The type species is Pararoseomonas rosea.
Description of Pararoseomonas rosea comb. nov.

Description of Pararoseomonas vinacea comb. nov.
Pararoseomonas vinacea (vi.na'ce.a. L. fem. adj. vinacea, of or belonging to wine or to the grape, referring to the colony color).
Basonym: Roseomonas vinacea (Zhang et al., 2008) The description is same as that given for Roseomonas vinacea by Zhang et al. (2008). The accession numbers for the 16S rRNA gene and genome sequences are MT760143 and BO93DRAFT, respectively. The type strain is CPCC 100056 T (=KCTC 22045 T = CCM 7468 T ). Basonym: Roseomonas nepalensis (Chaudhary and Kim, 2017) The description is the same as the one given for Roseomonas nepalensis by Chaudhary and Kim (2017). The accession number for the 16S rRNA gene sequence is KX129819. The type strain is G-3-5 T (=JCM 31470 T = KACC 18908 T ).

Basonym: Roseomonas aeriglobus (Lee and Jeon, 2018)
The description of Pararoseomonas aeriglobus is the same as that of Roseomonas aeriglobus as given by Lee and Jeon (2018). The accession number for the 16S rRNA gene sequence is KY864922. The type strain is KER25-12 T (=KACC 19282 T = JCM 32049 T ).
Basonym: Roseomonas aerilata (Yoo et al., 2008) The description is the same as that of Roseomonas aerilata as given by Yoo et al. (2008). Accession numbers of the 16S rRNA gene and genome sequences are EF661571 and JONP00000000, respectively. The type strain is 5420S-30 T (=KACC 12521 T = DSM 19363 T ).
Basonym: Roseomonas radiodurans  The description is the same as that given by Kim et al. (2018) for Roseomonas radiodurans. The accession number for the 16S rRNA gene sequence is KY887689. The type strain is 17Sr1-1 T (=KCTC 52899 T = NBRC 112872 T ).
Pararoseomonas pecuniae (pe.cu'ni.ae. L. gen. n. pecuniae, of/ from money or a coin, referring to the source of isolation of the type strain).
Basonym: Roseomonas pecuniae (Lopes et al., 2011) The description given is the same as that of Roseomonas pecuniae as given by Lopes et al. (2011). GU168019 is the accession number for the 16S rRNA gene sequence. The accession number for the genome sequence is JACIJD00000000. The type strain is N75 T (=LMG 25481 T = CIP 110074 T ).
Pararoseomonas harenae (ha.re'nae. L. gen. fem. n. harenae, of sand, referring to the isolation of the type strain from desert sand).
Basonym: Roseomonas harenae  The description is the same as that of Roseomonas harenae as given by Deng et al. (2020). MN883396 and WWDL00000000 are the accession numbers for the 16S rRNA gene and genome sequences, respectively. The type strain is CPCC 101081 T (=KCTC 62852 T = NBRC 113512 T ).
Members of this genus are aerobic, Gram-negative, and coccoid-rod in shape. Motility varies within the genus. Oxidase positive and catalase variable. Polar lipids consist of diphosphatidylglycerol, phosphatidylglycerol, phosphatidylethanolamine, phosphatidylcholine, unidentified phospholipid, unidentified glycolipid, and unidentified aminolipid. Summed feature 3, C 16:1 ω 5cand C 16:0 are the major fatty acids. Members have a genomic size of 5.8-6.6 Mb and a genomic G+C content of 70-72 mol%. The genus delineation is based on the 16S rRNA gene, 92 core genes (phylogenomics), AAI, POCP values, and phenotypic and genomic features.
The type species is Falsiroseomonas stagni.
Basonym: Roseomonas stagni (Furuhata et al., 2008) The species description is the same as that of Roseomonas stagni as given by Furuhata et al. (2008). AB369258 is the accession number for the 16S rRNA gene sequence and FOSQ00000000 for the genome sequence. The type strain is HS-69 T (=DSM 19981 T = JCM 15034 T = KCTC 22213 T ).
Description of Falsiroseomonas bella comb. nov.
Basonym: Roseomonas bella  The description is the same as that given by Zhang et al. (2020) for Roseomonas bella. MG601506 is the accession number for the 16S rRNA gene sequence and QGNA00000000 for the genome sequence. The type strain is CQN31 T (=KCTC 62447 T = MCCC 1H00309 T ).
Basonym: Roseomonas wooponensis (Lee et al., 2015) The description is the same as that of Roseomonas wooponensis as given by Lee et al. (2015). KF619243 is the accession number for the 16S rRNA gene sequence. The type strain is WW53 T (=KCTC 32534 T = JCM 19527 T ).
Basonym: Roseomonas terricola  The description is the same as that given for Roseomonas terricola by Kim et al. (2017). FJ654263 is the accession number for the 16S rRNA gene sequence. The type strain is EM302 T (=KACC 13942 T = KCTC 42906 T = NBRC 111477 T ).
Basonym: Roseomonas selenitidurans (Hou et al., 2020) The description is the same as that given for Roseomonas selenitidurans by Hou et al. (2020). MT107132 is the accession number for the 16S rRNA gene sequence and JAAVNE00000000 for the genome sequence. The type strain is BU-1 T (=GDMCC 1.1776 T = KACC 21750 T ).
Basonym: Roseomonas frigidaquae (Kim et al., 2009) The description is the same as that given by Kim et al. (2009) for Roseomonas frigidaquae. EU210160 is the accession number for the 16S rRNA gene sequence and JAAVTX00000000 for the genome sequence. The type strain is CW67 T (=JCM 15073 T = KCTC 22211 T ).
Basonym: Roseomonas tokyonensis (Furuhata et al., 2014) The description is the same as the one given for Roseomonas tokyonensis by Furuhata et al. (2013). AB297501 is the gene accession number for the 16S rRNA gene sequence. The type strain is K-20 T (=JCM 14634 T = KCTC 32152 T ).
Basonym: Roseomonas riguiloci (Baik et al., 2012) The description is the same as that given by Baik et al. (2012) for Roseomonas riguiloci. HQ436503 is the accession number for the 16S rRNA gene sequence. The type strain is 03SU10-P T (=KCTC 23339 T = JCM 17520 T = DSM 29515 T ).
Description of Falsiroseomonas algicola comb. nov. Basonym: Roseomonas algicola (Kim et al., 2020) The description is the same as the one given by Kim et al. (2020) for Roseomonas algicola. MK342491 is the accession number for the 16S rRNA gene sequence and JAAIKB00000000 for the genome sequence. The type strain is PeD5 T (=JCM 33309 T = KACC 19925 T ). Basonym: Roseomonas arcticisoli (Kim M. C. et al., 2016) The description is the same as the one given by Kim M. C. et al. (2016)  Members of this genus are strictly aerobic, Gram-negative, and coccoid-rod shaped. Cells variable motile. Oxidase negative and catalase positive. Diphosphatidylglycerol, phosphatidylglycerol, phosphatidylethanolamine, and phosphatidylcholine are the major polar lipids. Summed feature 3, C 18:1 ω7c, C 18:1 2-OH, and C 16:0 are the major fatty acids. G+C% content of 68.6-73.1 (mol%). The genus delineation is based on the 16S rRNA gene-based phylogeny and phenotypic features.

Description of
The type species is Paeniroseomonas aquatica.
Description of Paeniroseomonas aquatica comb. nov.
Basonym: Roseomonas aquatica (Gallego et al., 2006) The description is the same as that given by Gallego et al. (2006) for Roseomonas aquatica. AM231587 is the accession number for the 16S rRNA gene sequence. The type strain is TR53 T (=CECT 7131 T = JCM 13556 T = DSM 19438 T ).
Basonym: Roseomonas fluminis (Ko et al., 2018) The description is the same as that of Roseomonas fluminis as given by Ko et al. (2018). KY649439 is the accession number for the 16S rRNA gene sequence. The type strain is D3 T (=JCM 31968 T = KACC 19269 T ).
Description of Plastoroseomonas gen. nov.
Members of the genus are aerobic, Gram-negative, and rod shaped. Cells are motile. Oxidase negative and catalase positive.
Diphosphatidylglycerol, phosphatidylglycerol, phosphatidylethanolamine, unidentified aminolipid, and unidentified lipid are the major polar lipids. Summed feature 3, C 18:1 2-OH and C 16:0 are the major fatty acids. Genomic size of members ranges from 4.4 to 7.2 Mb and G+C content is ∼70 mol%. The genus delineation is based on the 16S rRNA gene, 92 core genes (phylogenomic), AAI indices, POCP values, and genomic and phenotypic features.
The type species is Plastoroseomonas arctica.
Description of Plastoroseomonas arctica comb. nov.
Plastoroseomonas arctica (arc'ti.ca. L. fem. adj. arctica, northern, from the Arctic, referring to the site where the type strain was isolated).
Basonym: Roseomonas arctica (Qiu et al., 2016) The description is the same as that of Roseomonas arctica as given by Qiu et al. (2016). KJ647399 and JAAEDH000000000 are the accession numbers for the 16S rRNA gene and genome sequences, respectively. The type strain is M6-79 T (=CCTCC AB 2013101 T = LMG 28251 T ).
Basonym: Roseomonas hellenica (Rat et al., 2021) The description is the same as that of Roseomonas hellenica as given by Rat et al. (2021). MN647549 and JAAGBB000000000 are the accession numbers for the 16S rRNA gene and genome sequences, respectively. The type strain is R-73080 T (=LMG 31523 T = CECT 30032 T ).
Description of Pseudoroseomonas gen. nov.
Members of the genus are aerobic, Gram-negative, and coccoid to short rods in shape. Cells are non-motile. Oxidase positive and catalase positive. Diphosphatidylglycerol, phosphatidylglycerol, phosphatidylethanolamine, unidentified aminolipid, unidentified lipid, and unidentified phospholipid are the major polar lipids. Summed feature 3, C 18:1 2-OH and C 16:0 are the major fatty acids. Genomic size of members ranges from 4.2 to 6.4 Mb and G+C% content is 68.7-72.7%. The genus delineation is based on the 16S rRNA gene, 92 core genes (phylogenomic), AAI indices, POCP values, and phenotypic and genomic features.
The type species is Pseudoroseomonas cervicalis. Basonym: Roseomonas cervicalis (Rihs et al., 1998) The description is the same as that of Roseomonas cervicalis as given by Rihs et al. (1993). AF533353 is the accession number for the 16S rRNA gene sequence and ADVL00000000 for the genome sequence. The type strain is E7107 T (=ATCC 49957 T = CIP 104027 T ).

Description of
Description of Pseudoroseomonas suffusca comb. nov.
Basonym: Roseomonas suffusca (Subhash and Lee, 2017) The description is the same as that of Roseomonas suffusca as given by Subhash and Lee (2017). LT009497 is the accession number for the 16S rRNA gene sequence. The type strain is S1 T (=KEMB 563-465 T = JCM 31176 T ).
Description of Pseudoroseomonas rubra comb. nov.
Basonym: Roseomonas rubra (Subhash et al., 2016) The description is the same as that of Roseomonas rubra as given by Subhash et al. (2016). LT009499 is the accession number for the 16S rRNA gene sequence. The type strain is S5 T (=JCM 31177 T = KEMB 563-468 T ). Basonym: Roseomonas hibiscisoli (Yan et al., 2017) The description is the same as that of Roseomonas hibiscisoli as given by Yan et al. (2017). KX456186 is the accession number for the 16S rRNA gene sequence. The type strain is THG-N2.22 T (=KACC 18935 T = CCTCC AB 2016176 T ).

Description of
Description of Pseudoroseomonas rhizosphaerae comb. nov.
Basonym: Roseomonas rhizosphaerae  The description is the same as that of Roseomonas rhizosphaerae as given by Chen et al. (2014). KC904962 is the accession number for the 16S rRNA gene sequence and PDNU00000000 for the genome sequence. The type strain is YW11 T (= KACC 17225 T = CCTCC AB2013041 T ).
Basonym: Roseomonas aestuarii (Venkata Ramana et al., 2010) The description is the same as that of Roseomonas aestuarii as given by Venkata Ramana et al. (2010). AB682256 is the accession number for the 16S rRNA gene sequence. The type strain is JC17 T (=CCUG 57456 T = KCTC 22692 T = NBRC 105654 T ).
Basonym: Roseomonas aerofrigidensis (Hyeon and Jeon, 2017) The description is the same as that of Roseomonas aerofrigidensis as given by Hyeon and Jeon (2017). KY126356 is the accession number for the 16S rRNA gene sequence. The type strain is HC1 T (=JCM 31878 T = KACC 19097 T ).
Description of Pseudoroseomonas oryzae comb. nov.
Pseudoroseomonas oryzae (o.ry'zae. L. gen. fem. n. oryzae, of rice, pertaining to the isolation of the type strain from rice paddy soil).
Basonym: Roseomonas oryzae (Ramaprasad et al., 2015) The description is to the same as that of Roseomonas oryzae, as described by Ramaprasad et al. (2015). LN810637 is the accession number for the 16S rRNA gene sequence and VUKA00000000 for the genome sequence. The type strain is JC288 T (=KCTC 42542 T = LMG 28711 T ).
Basonym: Roseomonas vastitatis  The description is the same as that of Roseomonas vastitatis as given by Zhao et al. (2020). MK421542 is the accession number for the 16S rRNA gene sequence and QXGS00000000 for the genome sequence. The type strain is CPCC 101021 T (=J1A743 T = KCTC 62043 T ).
Basonym: Roseomonas globiformis (Fang et al., 2018) The description is to the same as that of Roseomonas globiformis as given by Fang et al. (2018). MG589944 is the accession number for the 16S rRNA gene sequence. The type strain is CPCC 100847 T (=KCTC 52094 T ). Basonym: Roseomonas wenyumeiae (Tian et al., 2019) The description is the same as that of Roseomonas wenyumeiae as given by Tian et al. (2019). MH974806 is the accession number for the 16S rRNA gene sequence and RFLX00000000 for the genome sequence. The type strain is Z23 T (=CGMCC 1.16540 T = DSM 106207 T ). Basonym: Roseomonas ludipueritiae (Kämpfer et al., 2003;Sánchez-Porro et al., 2009) The description is the same as that of Roseomonas ludipueritia as given by Kämpfer et al. (2003) and Sánchez-Porro et al. (2009). AJ488504 is the accession number for the 16S rRNA gene sequence and JACTUZ0000000 for the genome sequence. The type strain is 170-96 T (=CIP 107418 T = DSM 14915 T ).

Description of
Description of Pseudoroseomonas aerophila comb. nov.
Basonym: Roseomonas aerophila  The description is the same as that of Roseomonas aerophile as given by Kim et al. (2013). JX275860 is the accession number for the 16S rRNA gene sequence and JACTVA00000000 for the genome sequence. The type strain is 7515T-07 T (=KACC 16529 T = NBRC 108923 T ).
Basonym: Roseomonas musae (Nutaratat et al., 2017) The description is the same as that of Roseomonas musae as given by Nutaratat et al. (2013). AB594201 is the accession number for the 16S rRNA gene sequence. The type strain is PN1 T (=BCC 44863 T = NBRC 107870 T ).
Basonym: Roseomonas coralli (Li et al., 2021) The description is the same as that of Roseomonas coralli as given by Li et al. (2021). MN336179 is the accession number for the 16S rRNA gene sequence and SNVJ000000000 for the genome sequence. The type strain is M0104 T (=KCTC 62359 T = MCCC 1K03632 T ).
Basonym: Roseomonas deserti (Subhash and Lee, 2018) The description is the same as that of Roseomonas deserti as given by Subhash and Lee (2018). LT837512 is the accession number for the 16S rRNA gene sequence and MLCO00000000 for the genome sequence. The type strain is M3 T (=KEMB 2255-459 T = JCM 31275 T ).
Description of Neoroseomonas gen. nov.
Members are aerobic, Gram-negative, and coccoid to short rod in shape. Motility is variable within the members. Oxidase and catalase positive. Diphosphatidylglycerol, phosphatidylglycerol, phosphatidylcholine, phosphatidylethanolamine, unidentified glycolipid, unidentified aminolipid, unidentified lipid, and unidentified phospholipid are the major polar lipids. Summed feature 3, C 18:1 2-OH and C 16:0 are the major fatty acids. Genome size of members ranges from 4.7 to 6.3 Mb and G+C% content is 68.8-71.5%. The genus delineation is based on the 16S rRNA gene, 92 core genes (phylogenomic), AAI indices, POCP values, and phenotypic and genomic features.
The type species is Neoroseomonas lacus.
Description of Neoroseomonas lacus comb. nov.
Neoroseomonas lacus (la'cus. L. gen. masc. n. lacus, of a lake, indicating the site of isolation of this organism).
Basonym: Roseomonas lacus (Jiang et al., 2006) The description is the same as that of Roseomonas lacu as given by Jiang et al. (2006). AJ786000 is the accession number for the 16S rRNA gene sequence and BMKW00000000 for the genome sequence. The type strain is TH-G33 T (=CGMCC 1.3617 T = JCM 13283 T ).
Basonym: Roseomonas terrae (Yoon et al., 2007) The description is the same as that of Roseomonas terrae as given by Yoon et al. (2007). EF363716 and JAAEDI000000000 are the accession numbers for the 16S rRNA gene and genome sequences, respectively. The type strain is DS-48 T (=KCTC 12874 T = JCM 14592 T ).
Basonym: Roseomonas eburnea  The description is the same as that of Roseomonas eburnean as given by Wang et al. (2016). KF254767 and JAAEDL000000000 are the accession numbers for the 16S rRNA gene and genome sequences, respectively. The type strain is BUT-5 T (=CCTCC AB2013276 T = KACC 17166 T ).
Basonym: Roseomonas alkaliterrae (Dong et al., 2014) The description is the same as that of Roseomonas alkaliterrae as given by Dong et al. (2014). KF771274 is the accession number for the 16S rRNA gene sequence and JACIJE00000000 for the genome sequence. The type strain is YIM 78007 T (=BCRC 80644 T = JCM 19656 T = DSM 25895 T ).
Basonym: Roseomonas oryzicola (Chung et al., 2015) The description is the same as that of Roseomonas oryzicola as given by Chung et al. (2015). EU707562 is the accession number for the 16S rRNA gene sequence and JAAVUP00000000 for the genome sequence. The type strain is YC6724 T (=KCTC 22478 T = NBRC 109439 T ).
Description of Neoroseomonas soli comb. nov.
Basonym: Roseomonas soli (Kim and Ka, 2014) The description is the same as that of Roseomonas soli as given by Kim and Ka (2014). JN575264 and JAAEDM00000000 are the accession numbers for the 16S rRNA gene and genome sequences, respectively. The type strain is 5N26 T (=KACC 16376 T = NBRC 109097 T ).
Basonym: Roseomonas sediminicola (He et al., 2019) The description is the same as that of Roseomonas sediminicola as given by He et al. (2014). JQ349047 is the accession number for the 16S rRNA gene sequence. The type strain is FW-3 T (=KACC 16616 T = JCM 18210 T ).

DATA AVAILABILITY STATEMENT
The datasets presented in this study can be found in online repositories. The names of the repository/repositories and accession number(s) can be found in the article/ Supplementary Material.

AUTHOR CONTRIBUTIONS
AR, UJ, and NS designed the studies under the supervision of CS and CR. UJ, AR, NS, and GD performed the genomic and phylogenetic analysis. AR and UJ wrote the manuscript. CR and CS supervised the study, contributed to the text preparation, and revised the manuscript. All authors read and approved the final version of the manuscript.