Small-sample learning reveals propionylation in determining global protein homeostasis

Shui, Ke; Wang, Chenwei; Zhang, Xuedi; Ma, Shen; Li, Qinyu; Ning, Wanshan; Zhang, Weizhi; Chen, Miaomiao; Peng, Di; Hu, Hui; Fang, Zheng; Guo, An‐Yuan; Gao, Guanjun; Ye, Mingliang; Zhang, Luoying; Xue, Yu

doi:10.1038/s41467-023-38414-8

Cited by 9 publications

(4 citation statements)

References 89 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…This would further improve the prediction results and promote the biological learned by this framework to promote downstream tasks. Secondly, we intend to introduce the small-sample learning theory [ 47 ] to perform accurate and robust predictions of the fourth digit of the EC number in a data-driven and learnable way. Thirdly, the relationships between the amino acid motifs of enzymes detected in this framework and their enzymatic reactions will be explored.…”

Section: Discussionmentioning

confidence: 99%

ifDEEPre: large protein language-based deep learning enables interpretable and fast predictions of enzyme commission numbers

Tan,

Xiao,

Chen

et al. 2024

Briefings in Bioinformatics

View full text Add to dashboard Cite

Accurate understanding of the biological functions of enzymes is vital for various tasks in both pathologies and industrial biotechnology. However, the existing methods are usually not fast enough and lack explanations on the prediction results, which severely limits their real-world applications. Following our previous work, DEEPre, we propose a new interpretable and fast version (ifDEEPre) by designing novel self-guided attention and incorporating biological knowledge learned via large protein language models to accurately predict the commission numbers of enzymes and confirm their functions. Novel self-guided attention is designed to optimize the unique contributions of representations, automatically detecting key protein motifs to provide meaningful interpretations. Representations learned from raw protein sequences are strictly screened to improve the running speed of the framework, 50 times faster than DEEPre while requiring 12.89 times smaller storage space. Large language modules are incorporated to learn physical properties from hundreds of millions of proteins, extending biological knowledge of the whole network. Extensive experiments indicate that ifDEEPre outperforms all the current methods, achieving more than 14.22% larger F1-score on the NEW dataset. Furthermore, the trained ifDEEPre models accurately capture multi-level protein biological patterns and infer evolutionary trends of enzymes by taking only raw sequences without label information. Meanwhile, ifDEEPre predicts the evolutionary relationships between different yeast sub-species, which are highly consistent with the ground truth. Case studies indicate that ifDEEPre can detect key amino acid motifs, which have important implications for designing novel enzymes. A web server running ifDEEPre is available at https://proj.cse.cuhk.edu.hk/aihlab/ifdeepre/ to provide convenient services to the public. Meanwhile, ifDEEPre is freely available on GitHub at https://github.com/ml4bio/ifDEEPre/.

show abstract

Section: Discussionmentioning

confidence: 99%

ifDEEPre: large protein language-based deep learning enables interpretable and fast predictions of enzyme commission numbers

Tan,

Xiao,

Chen

et al. 2024

Briefings in Bioinformatics

View full text Add to dashboard Cite

show abstract

“…The functional diversity of the wild and the mutant type of GFAP protein molecule in presence of various functional group were investigated by various PTM software including Netphos 3.1 24 (Phosphorylation), NetGlyc 1.0, NetOGlyc 4.0 25 (glycosylation), GPS-SUMO 26 (sumoylation), GPS-SNO 27 (S-nitrosylation), GPS-PAIL (N-acetylation) 28 , PrePs 29 (Prenylation), GPS-Palm 30 (Palmitoylation), GPS-Uber (Ubiquitination) 31 , KprFunc 32 (Propinylation) and GPS-Msp 33 (Methylation). It is necessary for understanding the PTM, which eventually puts a transparent view in disease pathogenicity depending on various functionalities of the protein.…”

Section: Methodsmentioning

confidence: 99%

Alexander Disease: Screening andin silicofunctional characterization of variants on candidate gene

Malakar,

Das,

Yadav

et al. 2024

Preprint

View full text Add to dashboard Cite

BackgroundAlexander disease is a rare neurological ailment caused by the degeneration of the white matter of the brain accompanied with the formation of Rosenthal fibers, a unique cytoplasmic inclusion within astrocytes (non-nerve tissue) in the brain. Till to date only heterozygousde novomutation in Glial Fibrillary Acidic Protein (GFAP) gene is found to be associated for the disease phenotype.Methods and Resultthe aim of our study was to determine the genetic basis of an Indian-origin juvenile AxD patient with pathological symptoms of macrocephaly and psychomotor delay followed by regression, spastic parapresis and feeding difficulty. The patient was screened for mutation in candidate gene for AxD by Whole Exome Sequencing and the detected variants were further reconfirmed by Sanger sequencing. The clonicopathological feature were investigated and two heterozygous missence variants (c.983T>C and c. 626G>A) were indentify in GFAP gene of the patient. Familial screenings of both of these variants were predicted to be pathogenic or damaging by variousin silicomethods and prediction tools.ConclusionThis altered GFA protein get accumulated in cytoplasm of the astroglial cells, leading to formation of Rosenthal fibers, which impairs cell function. Future study, however, would be helpful to understand the functional mechanism of these variant in formation of Rosenthal fibers leading to AxD pathophysiology.

show abstract

“…For this research article, various PTM sites were investigated. Netphos 3.1 (Phosphorylation)[31], NetNGlyc 1.0, NetOGlyc 4.0 (Glycosylation)[32], GPS-SNO (S-nitrosylation)[33], GPS-SUMO (Sumoylation)[34], PrePs (Prenylation)[35], GPS-PAIL (N-acetylation)[36], GPS-Palm (Palmitoylation)[37], GPS-Uber (Ubiquitination)[38], GPS-MSP (Methylation)[39], KprFunc (Propionylation)[40].…”

Section: Methodsmentioning

confidence: 99%

In silicoStudy of the Uncertainly SignificantVCPVariants reveal major Structural, Functional fluctuations leading to potential Disease-Based Associations

Das,

Roychowdhury,

Das

2023

Preprint

View full text Add to dashboard Cite

Mutations in VCP are responsible for a plethora of fatal diseases like Amyotrophic Lateral Sclerosis, Inclusion body myopathy with Paget disease of bone and frontotemporal dementia type 1, Spastic paraplegia, Charcot-Marie-Tooth disease type 2Y, Dementia, and Osteitis Deformans to name a few. Till now, studies on disease phenotype relationship, structural, and functional modifications of the protein’s stability, phenotype, molecular dynamics, and post-translational modifications haven’t been performed. This in-silico study investigates the variants of VCP (R95C, R95G, A160P, R191P, R191Q) which have conflicting interests of pathogenicity and hence are often depreciated and lack data. In addition, this study screens the study cohort and the fatal diseases linked to all of these variants. Interestingly through various computational tools and disease-based population studies, it was found that these variants are often found in patients linked with fatal diseases. The protein-protein interaction showed UFD1 has a direct association with VCP. The physicochemical parameters showed that A160P had the highest fluctuations of all the variants. By VCP’s molecular dynamics simulations, its variation, deviation, compactness, flexibility, hydrogen bonds, and solvent accessibility were all comparatively impacted due to the changes caused by the variants. Box-plot and Principal component analysis of the MD simulations visualized the changes in the wild type of the protein. Research in wet labs and screening of patient cohorts is necessary to investigate further the diseases linked with these variants and will be valuable to establish their specific roles and can further be used as biomarkers for fatal rare diseases.

show abstract

Small-sample learning reveals propionylation in determining global protein homeostasis

Cited by 9 publications

References 89 publications

ifDEEPre: large protein language-based deep learning enables interpretable and fast predictions of enzyme commission numbers

ifDEEPre: large protein language-based deep learning enables interpretable and fast predictions of enzyme commission numbers

Alexander Disease: Screening andin silicofunctional characterization of variants on candidate gene

In silicoStudy of the Uncertainly SignificantVCPVariants reveal major Structural, Functional fluctuations leading to potential Disease-Based Associations

Contact Info

Product

Resources

About