| Intrinsically disordered proteins have no stable three-dimensional structure in natural state,but can still perform their biological functions.In recent years,more and more intrinsically disordered proteins have been discovered and stored in relevant databases with the rapid development of protein structure determination technology.According to statistics,about 30% of proteins in eukaryotes contain disordered regions.Disordered regions of intrinsically disordered proteins have become research hotspots in recent years because of disordered regions are associated with many biological functions of intrinsically disordered proteins.It is generally believed in biology that the sequence of the protein determines the structure of the protein,and the structure of the protein determines the function of the protein.Since the protein sequence is translated from m RNA,the m RNA sequence information affects the structure of the protein to some extent.Most of the methods currently developed by researchers to predict the disordered regions of intrinsically disordered proteins are based on the protein level,and few computational methods identify disordered regions from the m RNA level.In view of this,we proposed a computational method to identify disordered regions of intrinsically disordered proteins at the protein and m RNA levels.And by using support vector machine and Jackknife test,the overall prediction success rate of 83.17% was finally achieved.Intrinsically disordered proteins have stronger binding affinities to other ligands due to their flexible conformations and large spatial coverage.The current database of binding sites for intrinsically disordered proteins contains few entries and is not fully annotated.It is difficult for researchers to use them to study the mechanism of action of intrinsically disordered proteins in relation to other ligands and to develop computational methods to identify binding sites of intrinsically disordered proteins.In this part,an intrinsically disordered protein binding site database IDPs Bind was developed and constructed by obtaining information from protein-related databases and manually annotating information.IDPs Bind contains a total of 880 intrinsically disordered proteins.It is currently the largest and most complete database of intrinsically disordered protein binding sites.Users can access it through http://www.s-bioinformatics.cn/idpsbind/home.The interaction of proteins with DNA and RNA is related to various physiological processes in organisms,such as DNA transcription,repair and replication,gene expression and regulation,and protein post-translational modification.Recent genetic and proteomic studies have shown that protein-DNA and RNA interactions are also involved in many human diseases,such as neurological diseases,cancer,and cardiovascular diseases.In this part,a method for predicting DNA and RNA binding sites in disease-related proteins was developed by using the position-specific scoring matrix(PSSM),amino acid index,etc.as feature parameters,using sliding window feature processing method and ensemble learning.This performance outperforms that of other existing methods and this method can be accessed at http://www.s-bioinformatics.cn/epdrna. |
