The author would like to thank Tan Ke En and Dr. Cameron Bracken for their critical review of this manuscript. This work was supported by grants from Fundamental Research Grant Scheme (FRGS/1/2020/SKK0/UM/02/15) and University of Malaya High Impact Research Grant (UM.C/625/1/HIR/MOE/CHAN/02/07).

In this protocol, de-identified ribosomal RNA (rRNA)-depleted RNA-seq library datasets prepared from the influenza A virus-infected human macrophage cells were downloaded and used from the Gene Expression Omnibus (GEO) database. The entire bioinformatics pipeline from prediction to functional characterization of circRNAs is summarized in Figure 1. Each part of the pipeline is further explained in the sections below.
1. Preparation, download, and setup before...

To illustrate the utility of this protocol, RNA-seq from influenza A virus-infected human macrophage cells was used as an example. CircRNAs functioning as potential miRNA sponges in host-pathogen interactions and their GO and KEGG functional enrichment within a host were investigated. Although there are a variety of circRNA tools available online, each of them is a standalone package that does not interact with one another. Here, we put together few of the tools that are required for circRNA prediction and quantification...

Host-pathogen interactions represent a complex interplay between the pathogens and host organisms, which triggers the hosts&#39; innate immune responses that eventually result in the removal of invading pathogens1,2. During&#160;pathogenic infections, a multitude of the host immune genes is regulated to inhibit the replication and release of pathogens. For example, common interferon-stimulated genes (ISGs) regulated upon pathogenic infections include ADAR1, IFIT1, IFIT2, IFIT3, ISG20, RIG-I, and OASL3,4. Besides protein-coding genes, studies have also reported that non-coding RNAs such as long non-coding RNAs (lncRNAs), microRNAs (miRNAs), and circular RNAs (circRNAs) also play a role and are regulated concurrently during pathogenic infections5,6,7. In contrast to protein-coding genes that mainly encode proteins as functional molecules, non-coding RNAs (ncRNAs) are known to function as regulators of genes at transcriptional and post-transcriptional levels. However, studies involving the participation of non-coding RNAs, particularly circRNAs, in regulating the hosts&#39; immune genes are not well reported compared to the protein-coding genes.
CircRNAs are widely characterized by their covalently closed continuous loop structure, which is generated through a non-canonical splicing process called back-splicing8. The process of back-splicing, unlike the splicing process of cognate linear RNAs, involves the ligation of the downstream donor site to the upstream acceptor site, forming a circular-shaped structure. Currently, three different back-splicing mechanisms for the biogenesis of circRNAs have been proposed. These are RNA binding protein (RBP) mediated circularization9,10, intron-pairing-driven circularization11, and lariat-driven circularization12,13,14. Given that circRNAs are connected end-to-end in a circular structure, they tend to be naturally resistant to normal exonuclease digestions and, thus, are considered to be more stable than their linear counterparts15. Another common characteristic exhibited by circRNAs includes the cell or tissue type-specific expression in hosts16.
As implied by their unique structure and cell or tissue-specific expression, circRNAs have been discovered to play important biological functions in cells. To date, one of the prominent functions of circRNAs is their role as microRNA (miRNA) sponges17,18. This regulatory role of circRNAs occurs through the complementary binding of circRNA nucleotides with the seed region of miRNAs. Such a circRNA-miRNA interaction inhibits the miRNAs&#39; normal regulatory functions on target mRNAs, thus regulating the expression of genes19,20. Additionally, circRNAs are also known to regulate gene expression by interacting with RNA binding proteins (RBPs) and forming RNA-protein complexes21. Although circRNAs are classified as non-coding RNAs, there is also evidence that circRNAs can act as templates for protein translation22,23,24.
Recently, circRNAs have been demonstrated to play pivotal roles in regulating the host-pathogen interactions, particularly between the hosts and viruses. Generally, host circRNAs are assumed to assist in regulating the host&#39;s immune responses to eliminate the invading pathogens. An example of&#160;circRNA that promotes host immune responses is circRNA_0082633, reported by Guo et al.25. This circRNA enhances type I interferon (IFN) signaling within A549 cells, which helps to suppress influenza virus replication25. Moreover, Qu et al.&#160;also reported a human intronic circRNA, called circRNA AIVR, that promotes immunity by regulating the expression of CREB-binding protein (CREBBP), a signal transducer of IFN-&#946;26,27. However, circRNAs that are known to promote the pathogenesis of disease upon infection also exist. For example, Yu et al. recently reported the role played by a circRNA spliced from the GATA zinc finger domain containing the 2A gene (circGATAD2A) in promoting the H1N1 virus replication through the inhibition of host cell autophagy28.
To effectively study circRNAs, a genome-wide circRNA prediction algorithm is usually implemented, followed by an in silico characterization of the predicted circRNA candidates before any functional studies can be carried out. Such a bioinformatics approach to predict and characterize circRNAs is less costly and more time efficient. It helps to refine the number of candidates to be functionally studied and could potentially lead to novel findings. Here, we provide a detailed bioinformatics-based protocol for the in silico identification, characterization and functional annotation of circRNAs during the host-pathogen interactions. The protocol includes the identification and quantification of circRNAs from RNA-sequencing datasets, annotation via circBase, and the characterization of the circRNA candidates in terms of circRNA types, number of overlapping genes, and predicted circRNA-miRNA interactions. This study also provides the functional annotation of the circRNA parental genes through Gene Ontology (GO) and the Kyoto Encyclopedia of Genes and Genomes (KEGG) enrichment analysis.

<table border="1" fo:keep-together.within-page="1" fo:keep-with-next.within-page="always">
	<tbody>
		<tr>
			<td>Bedtools</td>
			<td>GitHub</td>
			<td>https://github.com/arq5x/bedtools2/</td>
			<td>Referring to section 4.1.2. Needed for Circr.</td>
		</tr>
		<tr>
			<td>BWA</td>
			<td>Burrows-Wheeler Aligner</td>
			<td>http://bio-bwa.sourceforge.net/</td>
			<td>Referring to section 2.1.1 and 2.1.2. Needed to run CIRIquant, and to index the genome</td>
		</tr>
		<tr>
			<td>Circr</td>
			<td>GitHub</td>
			<td>https://github.com/bicciatolab/Circr</td>
			<td>Referring to section 4. Use to predict the miRNA binding sites</td>
		</tr>
		<tr>
			<td>CIRIquant</td>
			<td>GitHub</td>
			<td>https://github.com/bioinfo-biols/CIRIquant</td>
			<td>Referring to section 2.1.3. To predict circRNAs</td>
		</tr>
		<tr>
			<td>Clusterprofiler</td>
			<td>GitHub</td>
			<td>https://github.com/YuLab-SMU/clusterProfiler</td>
			<td>Referring to section 7. For GO and KEGG functional enrichment</td>
		</tr>
		<tr>
			<td>CPU</td>
			<td>Intel</td>
			<td>&#160;Intel(R) Xeon(R) CPU E5-2620 V2 @ 2.10 GHz&#160;&#160; Cores: 6-core CPU Memory: 65 GB Graphics card: NVIDIA GK107GL (QUADRO K2000)&#160;</td>
			<td>Specifications used to run this entire protocol.</td>
		</tr>
		<tr>
			<td>Cytoscape</td>
			<td>Cytoscape</td>
			<td>https://cytoscape.org/download.html</td>
			<td>Referring to section 5.2. Needed to plot ceRNA network</td>
		</tr>
		<tr>
			<td>FastQC</td>
			<td>Babraham Bioinformatics</td>
			<td>https://www.bioinformatics.babraham.ac.uk/projects/fastqc/</td>
			<td>Referring to section 1.2.1. Quality checking on Fastq files</td>
		</tr>
		<tr>
			<td>HISAT2</td>
			<td></td>
			<td>http://daehwankimlab.github.io/hisat2/</td>
			<td>Referring to section 2.1.1 and 2.1.2. Needed to run CIRIquant, and to index the genome</td>
		</tr>
		<tr>
			<td>Linux</td>
			<td>Ubuntu 20.04.5 LTS (Focal Fossa)</td>
			<td>https://releases.ubuntu.com/focal/</td>
			<td>Needed to run the entire protocol. Other Ubuntu versions may still be valid to carry out the protocol.</td>
		</tr>
		<tr>
			<td>miRanda</td>
			<td></td>
			<td>http://www.microrna.org/microrna/getDownloads.do</td>
			<td>Referring to section 4.1.2. Needed for Circr</td>
		</tr>
		<tr>
			<td>Pybedtools</td>
			<td>pybedtools 0.8.2</td>
			<td>https://pypi.org/project/pybedtools/</td>
			<td>Needed for BED file genomic manipulation</td>
		</tr>
		<tr>
			<td>Python</td>
			<td>Python 2.7 and 3.6 or abover</td>
			<td>https://www.python.org/downloads/</td>
			<td>To run necessary library modules</td>
		</tr>
		<tr>
			<td>R</td>
			<td>The Comprehensive R Archive Network</td>
			<td>https://cran.r-project.org/</td>
			<td>To manipulate dataframes</td>
		</tr>
		<tr>
			<td>RNAhybrid</td>
			<td>BiBiServ</td>
			<td>https://bibiserv.cebitec.uni-bielefeld.de/rnahybrid</td>
			<td>Referring to section 4.1.2. Needed for Circr</td>
		</tr>
		<tr>
			<td>RStudio</td>
			<td>RStudio</td>
			<td>https://www.rstudio.com/</td>
			<td>A workspace to run R</td>
		</tr>
		<tr>
			<td>samtools&#160;</td>
			<td>SAMtools</td>
			<td>http://www.htslib.org/</td>
			<td>Referring to section 2.1.2. Needed to run CIRIquant</td>
		</tr>
		<tr>
			<td>StringTie</td>
			<td>Johns Hopkins University: Center for Computational Biology</td>
			<td>http://ccb.jhu.edu/software/stringtie/index.shtml</td>
			<td>Referring to section 2.1.2. Needed to run CIRIquant</td>
		</tr>
		<tr>
			<td>TargetScan</td>
			<td>GitHub</td>
			<td>https://github.com/nsoranzo/targetscan</td>
			<td>Referring to section 4.1.2. Needed for Circr</td>
		</tr>
	</tbody>
</table>

in silico identification and characterization of circrnas during host-pathogen interactions

Circular RNAs (circRNAs) are a class of non-coding RNAs that are formed via back-splicing. These circRNAs are predominantly studied for their roles as regulators of various biological processes. Notably, emerging evidence demonstrates that host circRNAs can be differentially expressed (DE) upon infection with&#160;pathogens (e.g., influenza and coronaviruses), suggesting a role for circRNAs in regulating host innate immune responses. However, investigations on the role of circRNAs during pathogenic infections are limited by the knowledge and skills required to carry out the necessary bioinformatic analysis to identify DE circRNAs from RNA sequencing (RNA-seq) data. Bioinformatics prediction and identification of circRNAs is crucial before any verification, and functional studies using costly and time-consuming wet-lab techniques. To solve this issue, a step-by-step protocol of in silico prediction and characterization of circRNAs using RNA-seq data is provided in this manuscript. The protocol can be divided into four steps: 1) Prediction and quantification of DE circRNAs via the CIRIquant pipeline; 2) Annotation via circBase and characterization of DE circRNAs; 3) CircRNA-miRNA interaction prediction through Circr pipeline; 4) functional enrichment analysis of circRNA parental genes using Gene Ontology (GO) and Kyoto Encyclopedia of Genes and Genomes (KEGG). This pipeline will be useful in driving future in vitro and in vivo research to further unravel the role of circRNAs in host-pathogen interactions.

The protocol enlisted in the previous section was modified and configured to suit the Linux OS system. The main reason is that most module libraries and packages involved in the analysis of circRNAs can only work on the Linux platform. In this analysis, de-identified ribosomal RNA (rRNA)-depleted RNA-seq library datasets prepared from the Influenza A virus-infected human macrophage cells were downloaded from the GEO database42 and used to generate the representative results.
<p class="jove_con...

Watch this Scientific Journal Video about In Silico Identification and Characterization of circRNAs During Host-Pathogen Interactions at JoVE.com

In Silico Identification and Characterization of circRNAs During Host-Pathogen Interactions

The protocol submitted here explains the complete in silico pipeline needed to predict and functionally characterize circRNAs from RNA-sequencing transcriptome data studying host-pathogen interactions.

In Silico Identification and Characterization of circRNAs During Host-Pathogen Interactions

Circular RNAs (circRNAs) are a class of non-coding RNAs that are formed via back-splicing. These circRNAs are predominantly studied for their roles as ...

Circular RNAs play important regulatory roles in different biological processes. This protocol is suitable for beginners to carry out circular RNA analysis in the area of host and pathogen interactions. Here we put together a few of the tools create a streamlined protocol required for secure RNA prediction and quantification, secure RNA functional enrichment, secure RNA, micro-RNA interaction prediction and CCE RNA network constructions.This streamline protocol can be applied to clinical samples to identify certain candidates, diagnostic and prognostic values in a host and pathogen interaction setting. I expect those who have no prior programming knowledge to struggle doing the initial phase of this technique. Therefore, I would advise learning the basics of the programming languages used in this technique.I believe usually looking at how the programming language is applied to be more informative and easier to understand compared to reading alone. To begin open a Linux terminal and in the directory of the host's reference genome execute the commands bwa index and hisat2-build to index the genome. Prepare a yml configuration file containing the name of the file, the path of tools, the path to the downloaded reference files and the path to the index files.Specify the library type of the RNA sequence data and execute the Ciriquant tool using either the default or manual parameters. Prepare a text file with a list of data containing the IDs of the RNA sequence data, the path to the GTF files outputted by Ciriquant and the grouping of the RNA sequence data whether it is a control or a treated group. On the Linux terminal, run prep_Ciriquant with a prepared text file as an input.This run will generate a list of files. Prepare a second text file with a list of data containing the RNA sequence IDs and the path to their respective string tie output. The file layout must be similar to the previously prepared text file without the grouping column run.Run prepde. py with this text file as an input to generate the gene count matrix files. Execute Ciri_DE_Replicate with the library_info.csv, circRNA_BSJ. csv and gene_count_matrix. csv files as input to output the final circRNA_DE.tsv file. To filter and determine the number of differentially expressed, or DE, circRNAs, open the circRNA_DE. tsv file with R or any other spreadsheet software.Unzip and extract the contents of the CircR file after downloading it from the CircR GitHub page using the relevant software, such as WinRar or 7-Zip. Into a new directory where the analysis will be conducted. Then install the prerequisite software applications such as SAMTools, miRanda, RNAhybrid and Pybedtools before conducting the circRNA miRNA analysis.Index the reference genome file of the organism of interest using the SAMtools FAIDX command and prepare an input file consisting of the coordinates of the DE circRNAs of interest in a tab delimited bed file. Next, execute Circr. py using Python3.And as arguments specify the circRNA input file, the faster genome of the organism of interest, the genome version of the selected organism, the number of threads, and the name of the output file in the command line. Once the Circr analysis is complete, the program outputs a circRNA-miRNA interaction file in the CSV format. Prepare a tab delimited file containing the circRNAs of interest and their target miRNA.The first column consists of the circRNA name. The second column specifies the type of RNA from the first column. The third column is the target miRNA.And the fourth column specifies the type of RNA from the third column. To construct the ceRNA network map, open the Cytoscape software, navigate to file, import, network from file, select the prepared file and upload it. Press the style button to change the visual style of the network.Then press the arrow on the right side of fill color, choose type for the column, discrete mapping for the mapping type and select the color desired for each RNA type. After that navigate to shape to change the shape of the nodes and follow the steps shown earlier. For gene ontology and KEGG analysis of the parental gene of the circRNAs, ensure the cluster profiler and org.Hs.eg.db packages have been installed in our studio. Import the DE circRNA information into the R studio workspace. If the user wishes to convert the parental gene names to other formats such as the entrezid use a function such as bidder.Use the gene ID as an input and run the gene ontology and enrichment analysis using the enrichGO function within the cluster profile or package using default parameters. Finally, run the KEGG enrichment analysis using the gene ID as the input and the enrichKEGG function within the cluster profiler package. The bubble plot of gene ontology enrichment analysis of DE circRNA parental genes is shown in this figure.The gene ratio on the x axis is the number of genes in that input list associated with a given gene ontology term divided by the total number of genes in that term. The dot size in the plot is represented by the count value which is the number of genes in the input list associated with a given gene ontology term. The bigger the size of the dots, the larger the number of input genes associated with the term.The dots in the plot are color-coded based on the pvalue which is calculated by comparing the observed frequency of an annotation term with the frequency expected by chance. The enrichment is statistically significant and plotted on the bubble plot only if the pvalue is less than 0.01. Here, the top three enrichments for the biological processes include the ribonucleoprotein complex biogenesis, the response to the virus and the regulation of the response to a biotic stimulus.While for the molecular functions, only the catalytic activity acting on RNA and single-stranded RNA binding are statistically enriched. For the cellular components, only the retromer complex is statistically enriched. This representative image shows the KEGG enrichment analysis of the DE circRNA parental genes in a bubble plot.Only two KEGG terms were enriched in this case, the influenza A and the virallife cycle pathways. One of the most important things when attempting this procedure is to ensure the correct trait type of the RNA circ dataset you are using when running injury one. The bio-formatic pipeline provided here helps in predicting the potential secular RNAs and the functional annotations.However, well-led verification will still be needed to provide solid evidence. This protocol will allow researchers to discover secure RNA and their potential functional roles in the different codes and pathogen interactions, which they can further study.

in-silico-identification-characterization-circrnas-during-host

In Silico Identification and Characterization of circRNAs During Host-Pathogen Interactions

Abstract