Study On Some Key Problems In Exogenous Plant-derived MiRNA In Human Body

Recently,researchers have found plant mi RNAs in animal bodies and concluded that these mi RNAs come from the plant mi RNAs in foods.This striking founding has profound effects in many fields,such as,biology,medical,agriculture and so on.Then,the plant xenomi R hypothesis was founded.Unfortunately,this hypothesis was not accepted widely,and the central point of the controversy is whether the detected plant mi RNAs really come from animal bodies or from sample contamination and instrumental errors.No authoritative evidences are available till now.It is worth noting that the basic ideas are similar in both supportive or opposed researches,which is supposing plant xenomi R exist at first;then,design animal feeding experiments and setting control groups(Generally,feed some animal with foods containing different abundance of plant mi RNA);at last,measure the abundance of plant mi RNAs in animal body fluids/tissues and draw either the supportive or opposed conclusions.However,such experiments ignore the differences in animal species,the differences in the plantfoods and the difference in animal body fluids/tissues,which may be the reason causing the contradiction in experiments;In addition,the limited species and the number of samples make it difficult to separate plant mi RNAs from contaminations or instrument errors.When this kind of “top-down” biology experiments cannot obtain the generally acceptable results,bioinformatics adopts “down-top” idea,which starts from the data and digs the evidences for supporting or opposing the plant xenomi R hypothesis.However,some questions are the present bioinformatics research about plant xenomi R: at first,from the point of sample collection,the numbers of samples from animals are usually small and species of those are disorder.That may cause multiple factors and noises are hide in the data,which makes it difficult for analysis.Secondly,from the point of data analysis,the analyses are usually simple and uncomprehensive,which makes the results of analysis unreliable.Nowadays,a large number of small RNA sequencing data are stored in public accessible database and these data contains the evidences for deciding whether plant mi RNA existing in animal bodies.Meanwhile,these data are high-dimension,highnoisy and large-quantity.This paper will make use of these data and try to find strong evidences to support or deny the plant xenomi R hypothesis;In the case of plant xenomi Rs really existing,this paper will predict all the plant mi RNAs which are possible to enter human bodies according to the features in mi RNA sequences and predict their possible functions in human body by considering the secondary structures near the mi RNA targets.This paper avoided the errors brought by the small size of samples and could effectively distinguish plant xenomi Rs from the false-positive results made by contaminations or instrumental errors,which offers the basic guarantee for further researched about the mechanisms,functions and applications of plant xenomi Rs.The contents of this paper are as following.(1)Bioinformatics validation of plant xenomi R based on a large number of small RNA sequencing dataCollect small RNA sequencing data from healthy human body fluids/tissues and store them in local computer cluster;Combine tools of sequencing data analysis reasonably and build a pipeline for analyzing plant mi RNAs in human samples;Extract plant mi RNAs in human samples using computer cluster,and get their profiles;Collect small RNA sequencing data from microbiome samples and obtain their plant mi RNA profiles using the similar pipeline;Collect mi RNA expression profiles of multiple species of plants;Compare the plant mi RNA profiles from human samples and those from microbiome samples,plant samples using statistics and bioinformatics tools;Compare plant mi RNA profiles from different body fluids/tissues to test if plant mi RNA in human body is tissue-specific.(2)Prediction for the plant mi RNAs which are likely to enter human bodies based on mi RNA sequence featuresCollect the plant mi RNAs which can or cannot exist in human as positive and negative samples and collect features reasonably to form a feature matrix;Use random forest model to learn the samples and perform the precision validation;At last,using the best model,predict the plant mi RNA which can exist in human body in all plant mi RNAs.(3)Prediction for the possible function of plant mi RNA in human body combining the RNA secondary structure near the candidate targetsBased on heuristic method,this paper proposes a novel RNA secondary structure prediction method considering the RNA dynamic folding rules and implements the method with C++;Screen the plant mi RNAs targets in human genes for the first time using mi Randa and Target Scan,which are two kinds of targets prediction tools and the predicted targets compose candidate target set;Predict the secondary structures near the candidate targets using our novel method mentioned above,and screen the targets for the second time using the information of predicted structures;Perform the biological process and pathway analyses based on the selected targets.It is worth noting that this paper does not mean genetically modified crops could have any effects on human.Taken together,in order to prove disprove plant mi RNA hypothesis,this paper will start from a large number of human body fluids/tissues samples,combine both statistic and bioinformatic methods and find solid evidence for supporting or denying plant xenomi R hypothesis.Based on the above evidences,this paper will predict the mi RNAs which can exist in human body using machine learning methods.At last,considering the secondary structures near the candidate targets,this paper will build a plant mi RNA targets screening pipeline to obtain the high-probability target set for predicting their possible function in human body.This paper offers the basic guarantee for further researches about the mechanisms,functions and applications of plant xenomi Rs and has important theoretical and practical significance.