Protein-DNA Binding Residue Prediction via Bagging Strategy and Sequence-based Cube-Format Feature

Author(s):  
Jun Hu ◽  
Yan-Song Bai ◽  
Lin-Lin Zheng ◽  
Ning-Xin Jia ◽  
Dong-Jun Yu ◽  
...  
2018 ◽  
Vol 19 (S19) ◽  
Author(s):  
Lei Deng ◽  
Juan Pan ◽  
Xiaojie Xu ◽  
Wenyi Yang ◽  
Chuyao Liu ◽  
...  

2020 ◽  
Vol 29 (9) ◽  
pp. 1417-1425 ◽  
Author(s):  
Claire E L Smith ◽  
Laura L E Whitehouse ◽  
James A Poulter ◽  
Laura Wilkinson Hewitt ◽  
Fatima Nadat ◽  
...  

Abstract Amelogenesis is the process of enamel formation. For amelogenesis to proceed, the cells of the inner enamel epithelium (IEE) must first proliferate and then differentiate into the enamel-producing ameloblasts. Amelogenesis imperfecta (AI) is a heterogeneous group of genetic conditions that result in defective or absent tooth enamel. We identified a 2 bp variant c.817_818GC>AA in SP6, the gene encoding the SP6 transcription factor, in a Caucasian family with autosomal dominant hypoplastic AI. The resulting missense protein change, p.(Ala273Lys), is predicted to alter a DNA-binding residue in the first of three zinc fingers. SP6 has been shown to be crucial to both proliferation of the IEE and to its differentiation into ameloblasts. SP6 has also been implicated as an AI candidate gene through its study in rodent models. We investigated the effect of the missense variant in SP6 (p.(Ala273Lys)) using surface plasmon resonance protein-DNA binding studies. We identified a potential SP6 binding motif in the AMBN proximal promoter sequence and showed that wild-type (WT) SP6 binds more strongly to it than the mutant protein. We hypothesize that SP6 variants may be a very rare cause of AI due to the critical roles of SP6 in development and that the relatively mild effect of the missense variant identified in this study is sufficient to affect amelogenesis causing AI, but not so severe as to be incompatible with life. We suggest that current AI cohorts, both with autosomal recessive and dominant disease, be screened for SP6 variants.


2016 ◽  
Vol 12 (12) ◽  
pp. 3643-3650 ◽  
Author(s):  
H. Chai ◽  
J. Zhang ◽  
G. Yang ◽  
Z. Ma

A dynamic query-driven learning scheme helps to make more use of proteins with known structure and functions.


2018 ◽  
Vol 16 (03) ◽  
pp. 1840009 ◽  
Author(s):  
Xin Ma ◽  
Jing Guo ◽  
Xiao Sun

The identification of microRNA (miRNA)-binding protein residues significantly impacts several research areas, including gene regulation and expression. We propose a method, PmiRBR, which combines a novel hybrid feature with the Laplacian support vector machine (LapSVM) algorithm to predict miRNA-binding residues in protein sequences. The hybrid feature is constituted by secondary structure, conservation scores, and a novel feature, which includes evolutionary information combined with the physicochemical properties of amino acids. Performance comparisons of the various features indicate that our novel feature contributes the most to prediction improvement. Our results demonstrate that PmiRBR can achieve 85.96% overall accuracy, with 43.89% sensitivity and 90.56% specificity. PmiRBR significantly outperforms other approaches at miRNA-binding residue prediction.


Sign in / Sign up

Export Citation Format

Share Document