DeepNOG: fast and accurate protein orthologous group assignment

Feldbauer, Roman; Gosch, Lukas; Lüftinger, Lukas; Hyden, Patrick; Flexer, Arthur; Rattei, Thomas

doi:10.1093/bioinformatics/btaa1051

Cited by 21 publications

(14 citation statements)

References 33 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…In particular, we used a very low threshold on redundancy (<= 20% sequence identity). This is a much stricter threshold than applied in related studies on protein sequence classification such as DeepFam (35) and DeepNOG (36) which did not apply any redundancy removal filters, and Bepler and Berger’s study (33) that used a sequence identity filter of 40%, but CATHe achieved a comparable performance.…”

Section: Discussionmentioning

confidence: 99%

“…DeepFam attained a prediction accuracy of 97.17% on the family level on the GPCR dataset, and 95.4% on the COG dataset (with protein families having at least 500 sequences). DeepNOG (36) employed a method similar to DeepFam to classify sequences from COG and eggNOG5 databases. Furthermore, deep learning techniques have been used not just for family-level classification, but also for fold recognition, in FoldHSphere (37).…”

Section: Introductionmentioning

confidence: 99%

See 1 more Smart Citation

CATHe: Detection of remote homologues for CATH superfamilies using embeddings from protein language models

Nallapareddy

Bordin

Sillitoe

et al. 2022

Preprint

View full text Add to dashboard Cite

CATH is a protein domain classification resource that combines an automated workflow of structure and sequence comparison alongside expert manual curation to construct a hierarchical classification of evolutionary and structural relationships. The aim of this study was to develop algorithms for detecting remote homologues that might be missed by state-of-the-art HMM-based approaches. The proposed algorithm for this task (CATHe) combines a neural network with sequence representations obtained from protein language models. The employed dataset consisted of remote homologues that had less than 20% sequence identity. The CATHe models trained on 1773 largest, and 50 largest CATH superfamilies had an accuracy of 85.6+-0.4, and 98.15+-0.30 respectively. To examine whether CATHe was able to detect more remote homologues than HMM-based approaches, we employed a dataset consisting of protein regions that had annotations in Pfam, but not in CATH. For this experiment, we used highly reliable CATHe predictions (expected error rate <0.5%), which provided CATH annotations for 4.62 million Pfam domains. For a subset of these domains from homo sapiens, we structurally validated 90.86% of the predictions by comparing their corresponding AlphaFold structures with experimental structures from the CATHe predicted superfamilies.

show abstract

Section: Discussionmentioning

confidence: 99%

Section: Introductionmentioning

confidence: 99%

CATHe: Detection of remote homologues for CATH superfamilies using embeddings from protein language models

Nallapareddy

Bordin

Sillitoe

et al. 2022

Preprint

View full text Add to dashboard Cite

show abstract

“…CSBFinder-S (v0.6.3) (92) was used with the default settings to find the gmk-rpoZ synteny across 23,517 fully sequenced bacterial genomes downloaded from the NCBI genome database. DeepNOG (v1.2.3) (93) was run using the default setting to obtain the COG (Clusters of Orthologous Genes) ID for each gene. Strand information was obtained from the corresponding genomic.gff file for every genome downloaded.…”

Section: Methodsmentioning

confidence: 99%

Identification and structural modeling of the chlamydial RNA polymerase omega subunit

Cheng

Ghatak

et al. 2022

Preprint

View full text Add to dashboard Cite

Gene transcription in bacteria is carried out by the multisubunit RNA polymerase (RNAP), which is composed of a catalytic core enzyme and a promoter-recognizing σ factor. RNAP core enzyme comprises two α subunits, one β subunit, one β’ s subunit, and one ω (omega) subunit. Across multiple bacterial taxa, the RNAP ω subunit plays critical roles in the assembly of RNAP core enzyme and in other cellular functions, including regulation of bacterial growth, stress response, and biofilm formation. However, for several intracellular bacterium, including the obligate intracellular bacterium Chlamydia, no RNAP ω subunit previously has been identified. Here, we report the identification of Chlamydia trachomatis hypothetical protein CTL0286 as the chlamydial RNAP ω ortholog, based on sequence, synteny, and AlphaFold and AlphaFold-Multimer three-dimensional-structure predictions. We conclude that CTL0286 functions as the previously missing chlamydial ω ortholog. Extensions of our analysis indicate that all obligate intracellular bacteria have ω orthologs.IMPORTANCEChlamydiae are common mammalian pathogens. Chlamydiae have a unique developmental cycle characterized with an infectious but nondividing elementary body (EB), which can temporarily survive outside host cells, and a noninfectious reticulate body (RB), which replicates only intracellularly. Chlamydial development inside host cells can be arrested during persistence in response to adverse environmental conditions. Transcription plays a central role in the progression of the chlamydial developmental cycle as well as entry into and recovery from persistence. The identification of the elusive ω subunit of chlamydial RNAP makes possible future study of its regulatory roles in gene expression during chlamydial growth, development, and stress responses. This discovery also paves the way to prepare and study the intact chlamydial RNAP and its interactions with inhibitors in vitro.

show abstract

“…A Clusters of Orthologous Groups of proteins (COG) annotation analysis was performed using HMMER [52]. Gene Ontology (GO) functional enrichment and Kyoto Encyclopedia of Genes and Genome (KEGG) pathway analysis were carried out by Goatools (https://github.com/tanghaibao/Goatools, accessed on 7 June 2022) and KOBAS (http://kobas.cbi.pku.edu.cn/home.do, accessed on 7 June 2022) at a p value ≤ 0.05 or corrected p value ≤ 0.05 [53,54].…”

Section: Functional Annotation and Enrichmentmentioning

confidence: 99%

Exploration of Mycotoxin Accumulation and Transcriptomes of Different Wheat Cultivars during Fusarium graminearum Infection

Yan

et al. 2022

Toxins

View full text Add to dashboard Cite

Fusarium graminearum is one of the most devastating diseases of wheat worldwide, and can cause Fusarium head blight (FHB). F. graminearum infection and mycotoxin production mainly present in wheat and can be influenced by environmental factors and wheat cultivars. The objectives of this study were to examine the effect of wheat cultivars and interacting conditions of temperature and water activity (aw) on mycotoxin production by two strains of F. graminearum and investigate the response mechanisms of different wheat cultivars to F. graminearum infection. In this regard, six cultivars of wheat spikes under field conditions and three cultivars of post-harvest wheat grains under three different temperature conditions combined with five water activity (aw) conditions were used for F. graminearum infection in our studies. Liquid chromatography tandem mass spectrometry (LC–MS/MS) analysis showed significant differences in the concentration of Fusarium mycotoxins deoxynivalenol (DON) and its derivative deoxynivalenol-3-glucoside (D3G) resulting from wheat cultivars and environmental factors. Transcriptome profiles of wheat infected with F. graminearum revealed the lower expression of disease defense-factor-related genes, such as mitogen-activated protein kinases (MAPK)-encoding genes and hypersensitivity response (HR)-related genes of infected Annong 0711 grains compared with infected Sumai 3 grains. These findings demonstrated the optimal temperature and air humidity resulting in mycotoxin accumulation, which will be beneficial in determining the conditions of the relative level of risk of contamination with FHB and mycotoxins. More importantly, our transcriptome profiling illustrated differences at the molecular level between wheat cultivars with different FHB resistances, which will lay the foundation for further research on mycotoxin biosynthesis of F. graminearum and regulatory mechanisms of wheat to F. graminearum.

show abstract

DeepNOG: fast and accurate protein orthologous group assignment

Cited by 21 publications

References 33 publications

CATHe: Detection of remote homologues for CATH superfamilies using embeddings from protein language models

CATHe: Detection of remote homologues for CATH superfamilies using embeddings from protein language models

Identification and structural modeling of the chlamydial RNA polymerase omega subunit

Exploration of Mycotoxin Accumulation and Transcriptomes of Different Wheat Cultivars during Fusarium graminearum Infection

Contact Info

Product

Resources

About