Computing Identification And Bioninformatics Analyses Of Lysine Post-translational Modification Sites

Proteome is an important part of the life science research in the post-gene era.All cellular processes are tightly regulated via post-translational modifications(PTMs),which are an important mechanism for the regulation of protein function.As a difficult but hot topic,PTMs have attracted much attention.The identification and analyses of PTM sites are of great significance in understanding the function and function of substrate proteins together with the related-diseases.Comparing with the expensive and laborious experimental research,computational prediction could provide an accurate and effective approach to the identification of PTMs sites.And the related algorithms for the collection and statistical analyses of big data have attracted much attention.In this study,three computing identification tools were established for lysine malonylation,propionylation,and crotonylation which have rapid development in recent years.In addition,a series of proteomics analyses were carried out on seven PTM-types and their in-situ crosstalk.The main contents are as follows:1.Computing identification of specific-species lysine malonylaiton via enhanced characteristic strategy.Protein malonylation is a novel PTM which orchestrates a variety of biological processes and is related with type ? diabetes.Data analyses demonstrated that different organisms were preferentially involved in different biological processes and pathways.Meanwhile,unique sequence preferences were observed for each organism.A novel online prediction tool,called MaloPred(http://bioinfo.ncu.edu.cn/ MaloPred.aspx),which could predict malonylation for three species including Homo sapiens,Mus Musculus,and Escherichia coli,was developed by integrating and optimizing various informative features including sequence information,physicochemical properities and evolution information via enhanced characteristic strategy.Cross-validation and independent test demonstrated that the multiple features optimization could higly improve the performance comparing with the single type of feature.The satisfying results suggested that MaloPred could provide more instructive guidance for further experimental investigation of protein malonylation.2.Computing prediction and functional analysis of prokaryotic propionylation.As a novel PTM,lysine propionylation has attracted much attention in recent years.Although lots of propionylation sites have been identified in the past 10 years,the majority of lysine-propionylated substrates and their role in pathological physiology still remain largely unknown.By gathering various databases and literature,experimental prokaryotic propionylation data were collated to be trained in a support vector machine with various features via a three-step feature selection method.A novel online tool for seeking potential lysine propionylated sites(http://bioinfo.ncu.edu.cn/PropSeek.aspx)was constructed which could afford useful information for futher experimental validation and investigation.The cross-validation and independent test demonstrated that PropSeek was a stable and robust preidictor with a satisfying performance.Furthermore,high in-situ crosstalk with malonylation and acetylation suggested that propionylation could regulate protein function together with other acylations.Meanwhile,functional analyses implied a potential role of prokaryotic propionylation in protein synthesis and metabolism.3.Computing prediction of Lysine crotonylation before and after the Amino Acid Variant.As a novel and evolutionarily conserved PTM,lysine crotonylation has been identified in various diseases such as leukemia and cancers.The large-scale identification of lysine crotonylated sites afforded the opportunity for systematic anslyses the relationship between lysine crotonyaltion and related disease.By integrating different database and literature,a well-prepared up-to-data benchmark datasets were gathered together and trained with various features via a negative optimization by K nearest neighbor method.An online tool named as PredCrot(http://bioinfo.ncu.edu.cn/PredCrot.aspx)was constructed to predict lysine crotonylation before and after the amino acid variant.Systematical analyses demonstrated that the modified lysine were located in disordered regions,and had higher conversation scores comparing with non-modified lysine residue.By cross-referring the crotonylated proteins with the amino acid variant database,the disease-related modified lysine residues were identified,and the potential modified lysine with the amino acid mutation were predicted by PredCrot for further investigation,which could afford valuable information for the functional analyses of lysine crotonylation.4.Proteomic analysis of lysine modification and their in-situ crosstalk.Seven types of lysine modification including acetylation,glycation,malonylation,methylation,succinylation,SUMOylation and ubiquitination were combined and investigated together,and many particularities of modified lysine and proteins together with the in-situ crosstalk preference were revealed,including their enrichment Motifs,structural preference,uniform and biased functions and conservation character.The similarity of motifs and GO terms confirmed the in-situ cross preference for different PTM-types.Besides,modified lysine competitively suffering multiple PTM-types together with the disease related PTM-variation lysine was prone to be located in order regions with high conversation scores comparing with the single or non-modified ones.Moreover,the statistical results for the relationship between PTMs and amino acid variant demonstrated that acetylation and glycation were over-represented in human genetic variants.Furthermore,the investigation of transcription factors and their PTMs demonstrated that acetylated,SUMOylated,and ubiquited proteins were statistical enriched in cancer pathway.