Research On The Regulatory Mechanism And Function Prediction Of Alternative Splicing Events

Life science is dominating the 21 st century.The ultimate goal of life science is elucidating the regulation network of genes and proteins,which will lead to the understanding the principle of human disease and finally serve people.Alternative splicing is a process lie between transcription and translation,it enabled one gene encode several proteins and thus contributes to the protein diversity.The correct regulation of alternative splicing relies on the interaction between trans-acting elements and cis-acting elements,while dysfunction of either of the two can lead to serious disease.RNA binding proteins regulated alternative splicing process through binding on the ciselements of pre-mRNA and mutations in the RNA binding proteins can lead to serious disease.Recent works indicated that RBPs' regulatory function showed context dependent manners,binding in different regions could lead to opposite function.Here we integrated the eCLIP-seq and RNA-seq dataset from ENCODE project to reveal the regulatory patterns of 47 common RBPs.The results showed that most of the RBPs enhanced exon inclusion when binding in the exon region and repress exon inclusion when binding in the intron regions.Besides this,the results here also showed that although several hundreds of RBPs exist in the human genome,not all of them directly participate in the regulation of alternative splicing in K562 cell line.With the advent of NGS technology,once RNA-seq experiment can produce tens of thousands of alternative splicing events,while most of them are neutral,thus a model that can prioritize the alternative splicing events is needed.Here we used machine-learning technique and select dataset from HGMD and 1000 genome project as positive and negative dataset respectively,to try to find the intrinsic difference between the disease causing exons and the neutral exons.We found several features showed a strong difference between the two groups,and we further developed a tool called ExonImpact to make our model more useful for other researchers.Synonymous mutations didn't have direct impaction on protein function,rather they influence the protein expression levels by disrupting the cis-elements in the exon or in the flank intronic regions of the exon and influence the alternative splicing process.It was estimated that 90% of disease causing mutations lie outside of the coding regions.We used machine-learning techniques to try to evaluate the functional importance of synonymous mutation.The difference between our method and previous method is that our method takes protein features into account.The results support that the two groups have distinct patterns across various kinds of features.Lipopolysaccharide(LPS)is a gram-negative bacterial antigen that is capable of triggering a series of cellular responses in many cell types.A previous study suggests that LPS pre-condition improves the therapeutic efficacy of mesenchymal stem cells(MSCs)for repairing ischemic and injured tissues.In this study,we systematically evaluated the effects of LPS treatment in the genome-wide splicing pattern changes in MSCs of mouse,by comparing the transcriptome sequencing data in control and LPS treated samples.Overall,we identified 197 exons whose splicing patterns differed between the control and LPS-treated MSCs.Functional analysis of the alternatively spliced genes demonstrated significant enrichment of phosphoproteins,zinc finger proteins and protein acetylation.Additional bioinformatics analysis suggests that LPS-induced alternative spliced exons can have major effects on protein functions by interfering with key protein function domains,protein-protein interactions and post-translational modifications.Although it is still to be determined whether such modification improves MSC therapeutic efficacy,our characterizations would provide greater understanding of the mechanisms and clinical usage of the promising MSC therapies.