This article has Open Peer Review reports available.
RES-Scanner: a software package for genome-wide identification of RNA-editing sites
- Zongji Wang†1, 2,
- Jinmin Lian†2,
- Qiye Li†2, 3Email author,
- Pei Zhang2,
- Yang Zhou2,
- Xiaoyu Zhan2, 4 and
- Guojie Zhang2, 5Email author
© The Author(s). 2016
Received: 25 July 2015
Accepted: 4 August 2016
Published: 18 August 2016
High-throughput sequencing (HTS) provides a powerful solution for the genome-wide identification of RNA-editing sites. However, it remains a great challenge to distinguish RNA-editing sites from genetic variants and technical artifacts caused by sequencing or read-mapping errors.
Here we present RES-Scanner, a flexible and efficient software package that detects and annotates RNA-editing sites using matching RNA-seq and DNA-seq data from the same individuals or samples. RES-Scanner allows the use of both raw HTS reads and pre-aligned reads in BAM format as inputs. When inputs are HTS reads, RES-Scanner can invoke the BWA mapper to align reads to the reference genome automatically. To rigorously identify potential false positives resulting from genetic variants, we have equipped RES-Scanner with sophisticated statistical models to infer the reliability of homozygous genotypes called from DNA-seq data. These models are applicable to samples from either single individuals or a pool of multiple individuals if the ploidy information is known. In addition, RES-Scanner implements statistical tests to distinguish genuine RNA-editing sites from sequencing errors, and provides a series of sophisticated filtering options to remove false positives resulting from mapping errors. Finally, RES-Scanner can improve the completeness and accuracy of editing site identification when the data of multiple samples are available.
RES-Scanner, as a software package written in the Perl programming language, provides a comprehensive solution that addresses read mapping, homozygous genotype calling, de novo RNA-editing site identification and annotation for any species with matching RNA-seq and DNA-seq data. The package is freely available.
RNA editing is a post-transcriptional-processing mechanism, which alters RNA sequences by insertion, deletion or modification of specific nucleotides so that the information in the mature RNA differs from that defined in the genome [1, 2]. In metazoa, the vast majority of RNA-editing events involve the deamination of adenosine (A) to inosine (I), which is catalyzed by a family of adenosine deaminases that act on RNA (ADARs) . As inosine is recognized as guanosine (G) by other molecular machines in vivo, A-to-I editing provides a potential mechanism for diversifying the transcriptomes by recoding amino acids , changing messenger RNA (mRNA) splicing sites , editing microRNA (miRNA) sequences  or changing miRNA target sites in mRNA . Other types of RNA-editing events (e.g. C-to-U, U-to-C and G-to-A editing) are also documented [6–9], but considered to be rare in metazoa. RNA editing appears to frequently target transcripts that encode proteins involved in fast neuronal signaling . The dysregulation of RNA editing results in behavioral defects in Caenorhabditis elegans  and Drosophila melanogaster  and is associated with a variety of neurological diseases and cancers in humans [13, 14].
High-throughput RNA sequencing has enabled genome-wide identification of RNA-editing sites in any species. However, distinguishing RNA-editing events from genetic variants and technical artifacts caused by sequencing or read-mapping errors is still a challenge [15–18]. Much progress has been achieved in recent years in RNA-editing study at the genomic scale, resulting in the discovery of thousands to millions of RNA-editing sites in humans [19–24], rhesus macaques , mice , fruit flies , ants  and nematodes . At the same time, several methodologies have been proposed to accurately identify RNA-editing sites with matching RNA-seq and DNA-seq data or with RNA-seq data alone [21–24]. However, published software packages devoted to this aim are scarce, especially for de novo identification of RNA-editing sites in non-model species.
REDItools was the first published software package for genome-wide RNA-editing site identification. It uses pre-aligned HTS reads in BAM format as inputs and implements a variety of filters to remove potential false positives . However, it does not implement statistical models for determining homozygous genomic sites from DNA-seq data or for distinguishing real RNA-editing events from sequencing errors. GIREMI is a more recently released software package, which was developed to detect RNA-editing sites from RNA-seq data alone on the basis of allelic linkage and generalized linear models . However, a comprehensive single-nucleotide polymorphism (SNP) dataset of the studied species must be available for GIREMI to estimate the reference mutual information (MI) distribution, which limits its application in non-model species that do not have such information. Thus far, an automatic tool that integrates read mapping, homozygous genotype calling, RNA-editing site identification and annotation for both model and non-model species is still lacking.
Here we introduce RES-Scanner (RES: RNA-editing site), a flexible and efficient software package written in the Perl programming language, which has been developed for genome-wide identification and annotation of RNA-editing sites. RES-Scanner is designed to address HTS read mapping, homozygous genotype calling, and de novo RNA-editing site identification and annotation for any species with matching RNA-seq and DNA-seq data.
Methods and results
RNA/DNA-seq read mapping and filtering
While the inputs for RES-Scanner are single/paired-end reads in FASTQ format, RES-Scanner can invoke BWA, an effective and accurate short-read aligner , to map the reads against a combination of a reference genome and exonic sequences surrounding all known splicing junctions (hereafter called junction sequences). Following Ramaswami et al. [22, 23], we set the length of the junction sequences to be slightly shorter than the read length to avoid simultaneous hits to the reference genome and the junction sequences. For example, when reads are 90 bp in length, a region of 89 bp upstream and downstream is selected. After BWA alignment, the locations of reads mapped on junction sequences are converted to regular genomic locations before subsequent analysis. We chose BWA as the default aligner in RES-Scanner, because the applicability and accuracy of BWA in genome-wide RNA-editing site detection have been acknowledged by many published studies for a series of species, including human [18, 22, 23, 33], macaque [25, 34], mouse , chicken , fly , C. elegans  and ant . However, the default mapping strategy implemented in RES-Scanner cannot be optimal for every species, and so RES-Scanner can also accept as inputs pre-aligned DNA and RNA reads in BAM format from other aligners, such as Bowtie 2 , TopHat2 , GSNAP , and HISAT2 . Thus, users are afforded considerable freedom in their choice of aligner to map the raw reads.
In the next step, RES-Scanner only keeps those reads with unique alignment and with no suboptimal hits using Perl scripts, and discards potential PCR duplications (i.e. read pairs that mapped to identical genomic locations) except for the one with the highest mapping quality according to the ‘rmdup’ function of SAMtools . Given the higher error rate in Illumina sequencing at the ends of reads , the introduction of mismatches at the 5‘ read ends by random-hexamer priming during the first- and second-strand syntheses of RNA library construction , and the mapping errors at both the 5’ and 3’ ends resulting from the incorrect handling of insertions/deletions , false positives in RNA-editing sites are disproportionately increased at both ends of the reads. Thus, RES-Scanner clips the first and last few bases (six bases by default) of each aligned read before further analysis.
Homozygous genotype calling from DNA-seq data
Previous studies usually distinguished candidate RNA-editing events from genome-encoded variants based on frequency of the alternative allele (i.e. the allele that is not encoded in the reference genome) present in the DNA-seq data, followed by selection of an arbitrary threshold [19, 22, 25, 29]. However, this strategy may result in the under-estimation of heterozygous sites, especially when the sequencing depth is low or the sample is not diploid. Moreover, it does not provide a statistical measurement of uncertainty in the inference of homozygous genotypes.
RES-Scanner introduces a statistical model based on a Bayesian method (Bayesian model) to infer the genotype of each genomic site on the reference genome . This statistical model can handle samples from one individual or a pool of multiple individuals based on the ploidy information supplied by the users (e.g. the ploidy for a sample from a diploid individual is two and that of a pooled sample with two diploid individuals is four). In brief, for a genomic site on the reference genome, our model computes the probabilities of all the possible genotypes based on the nucleotide bases, along with their sequencing quality scores, mapped to this genomic site, and then reports the genotype with the highest posterior probability. For example, for a sample from a diploid individual, all ten possible genotypes (i.e. AA, TT, CC, GG, AT, AC, AG, CT, CG and GT) are examined, while for a sample obtained by pooling two diploid individuals, all 22 possible genotypes are examined (for details see Additional file 1: Supplementary text 1). Finally, only the genomic sites reporting homozygous genotypes with posterior probabilities exceeding 0.95 or 0.99 and supported by a sufficient number of reads (i.e. ≥ 10×) are kept for RNA-editing site determination. In practice, RES-Scanner only applies the Bayesian procedure to candidate RNA-editing sites that satisfy some basic criteria (see below) in order to save computing time.
RES-Scanner also provides two additional methods to estimate the reliability of a homozygous genotype: one is based on a binomial distribution model (Binomial model) and the other is based on the read depth and the frequency of the alternative allele (Frequency model) present in the DNA-seq data (for details see Additional file 1: Supplementary texts 2 and 3).
Identification of RNA-editing sites
By default, RES-Scanner requires that a candidate RNA-editing site in a given sample must be supported by at least three non-redundant RNA reads that have been mapped on overlapping but not identical positions, and the editing level (i.e. the percentage of all reads mapping on that position that support editing) of this site must be ≥ 5 %. Bases on RNA reads with a Phred quality score below 30 are discarded, restricting the upper limit of sequencing error for an RNA base to < 0.1 % for any candidate editing site. Users can modify the quality score cutoff according to their dataset. These candidate editing sites are then subjected to genotype analysis as described above, and only the sites showing robust homozygosity in genomic DNA are kept. RES-Scanner also allows users to supply a list of potential SNPs derived from other analyses or databases of the target species, and exclude candidate editing sites that overlap these SNPs. Candidate sites that show multiple editing types are also discarded as these positions may be associated with higher probabilities of mapping error.
To avoid potential false positives resulting from misalignment of reads onto very similar paralogous regions, RES-Scanner invokes BLAT , a BLAST-like alignment tool with a fundamentally different algorithm from most short-read aligners, to realign all the reads that support RNA editing (i.e. reads showing a mismatch to the reference). Then, a read is defined as a qualifying read if (1) the best hit of this read overlaps its original candidate site and (2) the second-best hit, if it exists, has a BLAT score of < 95 % of the best hit. Only candidate editing sites for which the proportion of qualifying reads in relation to all BLAT-realigned reads exceeds 50 % are kept.
In addition to these basic filters, and to avoid potential false positives resulting from mis-mapping of reads at splice junctions, RES-Scanner requires a candidate editing site to be supported by at least one RNA read in the middle of its length (e.g. from positions 23 to 68 of a 90-bp read), and discards intronic candidate sites that occur within six bases of a splice site. RES-Scanner also removes sites in homopolymer runs of five or more base pairs (e.g. AAAAA), given that homopolymers have higher rates of sequencing error . Finally, RES-Scanner also discards candidate editing sites with DNA read depths of more than twice the genome-wide peak or mean depth, as such sites are likely to be located in regions showing copy number variation (i.e. another form of highly similar paralogous region that is not fully present in the reference genome).
To further eliminate false positives due to sequencing errors, RES-Scanner performs statistical tests for all the candidate editing sites based on the binomial distribution B(k, n, p) , where p is set to be the upper limit of sequencing error for an RNA base as described above, n is equal to the total read depth of a given candidate site, and k denotes the number of reads supporting editing (for details see Additional file 1: Supplementary text 4). P-values are then adjusted by the Benjamini-Hochberg false discovery rate (FDR) , and only candidate sites with FDRs below a user-chosen cutoff (usually 0.01 or 0.05) are considered as true positives. Two examples from real data that show the positive contribution of binomial tests to reducing false positives resulting from sequencing error are shown in Additional file 1: Supplementary text 8.
Improvement of identification when multiple samples are available
Owing to the stringent filtering criteria of RES-Scanner in identifying RNA-editing sites for a single sample, a number of true positives may be missed due to their low editing level, insufficient sequencing depth or failure to satisfy other requirements. These editing sites can, however, be retrieved by RES-Scanner if multiple samples across individuals or tissues are available. RES-Scanner first combines all editing sites identified in each sample to obtain a comprehensive map of potentially editable positions in the genome of the target species. These positions are homozygous for DNA, located in unique genomic regions, not close to any splice sites and RNA-edited in at least one of the multiple samples. RES-Scanner then retrieves missed editing sites in each sample in these editable positions using the more liberal criteria of at least one RNA read supporting editing and an FDR below a user-chosen cutoff (usually 0.01 or 0.05). DNA data from multiple samples are also helpful in improving the accuracy of calling homozygous genomic sites. RES-Scanner will discard any editing site for which the genomic DNA is not homozygous in any one of the multiple DNA samples.
Annotation of RNA-editing sites
RES-Scanner can annotate the identified RNA-editing sites with a variety of genomic features, such as exon, intron, coding sequence (CDS), 5‘-untranslated region (5‘-UTR), 3‘-UTR and repeat, if the position files of these genomic features are provided (see Additional file 2 for format description). Moreover, for the editing sites targeting CDSs, RES-Scanner can further infer the codon and amino acid change after RNA editing.
The accuracy of RES-Scanner
To estimate the accuracy of RES-Scanner, we reanalyzed the leaf-cutting ant dataset from Li et al. (nine samples with matching DNA-seq and strand-specific RNA-seq data)  and the human GM12878 lymphoblastoid cell line dataset from the ENCODE project, which has been used as a benchmark by many studies (e.g. [20, 22, 24, 30]).
For the ant dataset, we ran RES-Scanner with default parameters, except for setting the ploidy to eight for the Bayesian model to determine homozygous genomic sites due to the special genetic background of the ant samples  (for details see Additional file 1: Supplementary text 5). We identified an average of 14,650 editing sites (range 10,282–20,234) per sample, with about 95 % representing A-to-I editing (Additional file 1: Table S1; Additional file 3). We then estimated the FDR using the TA-clonal sequencing data of 16 PCR amplicons generated by Li et al. . In total, we found 76 editing sites (71 A-to-I sites and five non-A-to-I sites) distributed over these PCR amplicons, with only three of them (two A-to-I sites and one non-A-to-I site) failing to be confirmed due to the absence of an observable editing signal in TA-clonal sequencing data (Additional file 3), representing a FDR of 4 % (3/76). Furthermore, we observed that the editing levels calculated by RES-Scanner were consistent with the levels obtained by TA-clonal sequencing (Pearson‘s r = 0.90 and p < 10−15; Additional file 1: Figure S1).
We next tested the performance of the three models (Bayesian, Binomial and Frequency) on homozygous genotype calling and RNA-editing site identification based on the data of sample L363 from Li et al. . In general, we found that the vast majority of homozygous genomic sites and RNA-editing sites identified were common to all three methods, indicating that they performed comparably well (Additional file 1: Figure S2). In practice, the Frequency model runs faster than the Binomial model, and the Binomial model faster than the Bayesian model. However, if the average depth of the DNA-seq data is either low (e.g. ≤ 10×) or particularly high (e.g. ≥ 50×), we recommend the use of the Bayesian or Binomial model to statistically estimate the genomic homozygosity for the candidate editing sites. This is expected to reduce false positives when DNA-seq depth is low and reduce false negatives when DNA-seq depth is high.
Performance of RES-Scanner compared with other methods applied to GM12878 human lymphoblastoid cell line data
Ramaswami et al. 
RES-Scanner (raw reads)
Taken together, these results indicate that RES-Scanner provides high accuracy in genome-wide RNA-editing site identification using matching DNA-seq and RNA-seq data, and it is capable of working on organisms with simple genomes like ants and complex genomes like humans.
The advantages of RES-Scanner relative to existing software
Comparison of the cumulative CPU times (hours) for RES-Scanner, REDItools and GIREMI in processing the human GM12878 dataset from pre-aligned reads to final editing sites
We also compared the performance of the three software packages (RES-Scanner, REDItools and GIREMI) on non-diploid samples from non-model species using the leaf-cutting ant dataset from Li et al. , in which each sample was a pool of multiple individuals from the same ant colony, representing samples with a ploidy of eight . GIREMI is unable to detect RNA-editing sites in the ant samples because its MI model is specifically designed for diploids . For REDItools and RES-Scanner, we used the same sets of DNA and RNA BAM files as inputs and used similar parameters for editing site identification (for details see Additional file 1:Supplementary text 10 and Table S8). However, for the parameter corresponding to the definition of homozygous genotypes from DNA-seq data in REDItools, we chose a series of cutoffs by limiting the maximal frequency of non-reference bases to be equal to 0, ≤ 0.02 and ≤ 0.05, respectively, as there was no prior knowledge about the optimal cutoff for samples with a ploidy of eight. We found that REDItools generally detected similar numbers of editing sites but with significantly lower A-to-I ratios when compared with RES-Scanner (~80 vs 94 %; Additional file 1: Table S8), implying that REDItools produced more false positives on the ant dataset given that non-A-to-I editing has been confirmed as rare in ants . Although the performance of these software packages on other datasets besides the ant data remains to be investigated, it is noteworthy that RES-Scanner implements statistical models (Bayesian and Binomial) to infer the homozygous genotypes from DNA-seq data instead of choosing arbitrary thresholds. These models make RES-Scanner applicable for distinguishing RNA-editing sites from genetic variants for samples with any ploidy number, greatly enhancing the value of RES-Scanner for other datasets.
Finally, RES-Scanner can provide an integrated, end-to-end solution that works from raw sequencing reads to final editing sites, greatly reducing the risk of incompatibility between read alignment outputs and the downstream editing site identification pipeline. RES-Scanner can automatically annotate identified RNA-editing sites with genomic features and deduce codon and amino acid changes after RNA editing, providing useful information for downstream analysis. Furthermore, the final result file output by RES-Scanner has integrated DNA and RNA information from all samples for each editing site, greatly facilitating downstream comparative analysis between different samples. Thus, we believe that RES-Scanner is also superior to other existing software packages in terms of ease of use.
It is equipped with rigorous statistical models (Bayesian and Binomial) to infer the reliability of homozygous genotypes derived from DNA-seq data. This approach is different from traditional SNP calling, which is aimed at ensuring the reliability of genetic polymorphism rather than homozygosity.
With these statistical models, RES-Scanner is capable of calling homozygous genotypes reliably from samples with any ploidy number, including samples from a pool of multiple individuals. This feature is especially useful because many species - or the target tissues of a species - are small in size and, in practice, pooling of multiple individuals is usually required to obtain enough biomass for sequencing.
RES-Scanner implements binomial tests to rigorously distinguish RNA-editing sites from sequencing errors by assigning a p-value to each RNA-editing candidate.
RES-Scanner provides a complete pipeline from raw sequencing reads to final editing sites, which should be especially valuable to users who have limited experience in bioinformatics or are working with non-model species with no prior knowledge of the optimal mapping strategy.
The application of our package to the leaf-cutting ant and GM12878 human datasets demonstrates that RES-Scanner provides high accuracy in identifying RNA-editing sites using matching DNA-seq and RNA-seq data. It should be noted that RES-Scanner is not species-specific: it is applicable to genome-wide identification of RNA-editing sites in any species with matching RNA-seq and DNA-seq data.
A detailed user manual for RES-Scanner is available in Additional file 2.
Availability and requirements
FDR, false discovery rate; HTS, high-throughput sequencing; MI, mutual information; SNP, single-nucleotide polymorphism; SNV, single-nucleotide variant
We thank Dr. Ernesto Picardi from the University of Bari for providing the GM12878 BAM files that were originally produced by Ramaswami et al.  and the RNA-editing sites identified with REDItools by Picardi et al. . We also thank Dr. Qing Zhang from the University of California for positively addressing the issues we encountered when using GIREMI. This research was supported by the National Natural Science Foundation of China (31271392 and 31501057), an EU Marie Curie International Incoming Fellowship (300837) for GZ, and the China National Genebank-Shenzhen.
Availability of data and materials
DNA-seq and RNA-seq data of the leaf-cutting ant Acromyrmex echinatior used in this study can be accessed from the National Center for Biotechnology Information (NCBI) Gene Expression Omnibus (GEO)  under accession number GSE51576. Genome assembly and annotation data of A. echinatior can be accessed in the GigaScience GigaDB database . Pre-aligned DNA-seq reads in BAM format for the GM12878 human cell line can be downloaded from NCBI  and the raw RNA-seq reads can be downloaded from the University of California, Santa Cruz genome website . Test data for RES-Scanner has been uploaded, together with the software package, to the project home page . Snapshots of the code and additional supporting information can be found in GigaDB .
QL, ZW and GZ designed the study. JL wrote the majority of scripts of the software package. PZ and ZW designed and wrote the Bayesian statistical core. QL wrote the Perl module for the binomial test. YZ and XZ provided testing of the software package on different platforms and performed the comparative analysis. QL, ZW, GZ and JL wrote the manuscript. All of the authors read and approved the final manuscript.
The authors declare that they have no competing interests.
Open AccessThis article is distributed under the terms of the Creative Commons Attribution 4.0 International License (http://creativecommons.org/licenses/by/4.0/), which permits unrestricted use, distribution, and reproduction in any medium, provided you give appropriate credit to the original author(s) and the source, provide a link to the Creative Commons license, and indicate if changes were made. The Creative Commons Public Domain Dedication waiver (http://creativecommons.org/publicdomain/zero/1.0/) applies to the data made available in this article, unless otherwise stated.
- Gott JM, Emeson RB. Functions and mechanisms of RNA editing. Annu Rev Genet. 2000;34:499–531.View ArticlePubMedGoogle Scholar
- Bass BL. RNA editing by adenosine deaminases that act on RNA. Annu Rev Biochem. 2002;71:817–46.View ArticlePubMedGoogle Scholar
- Laurencikiene J, Kallman AM, Fong N, Bentley DL, Ohman M. RNA editing and alternative splicing: the importance of co-transcriptional coordination. EMBO Rep. 2006;7:303–7.PubMedPubMed CentralGoogle Scholar
- Kawahara Y, Zinshteyn B, Sethupathy P, Iizasa H, Hatzigeorgiou AG, Nishikura K. Redirection of silencing targets by adenosine-to-inosine editing of miRNAs. Science. 2007;315:1137–40.View ArticlePubMedPubMed CentralGoogle Scholar
- Borchert GM, Gilmore BL, Spengler RM, Xing Y, Lanier W, Bhattacharya D, et al. Adenosine deamination in human transcripts generates novel microRNA binding sites. Hum Mol Genet. 2009;18:4801–7.View ArticlePubMedPubMed CentralGoogle Scholar
- Sharma PM, Bowman M, Madden SL, Rauscher 3rd FJ, Sukumar S. RNA editing in the Wilms‘ tumor susceptibility gene, WT1. Genes Dev. 1994;8:720–31.View ArticlePubMedGoogle Scholar
- Villegas J, Muller I, Arredondo J, Pinto R, Burzio LO. A putative RNA editing from U to C in a mouse mitochondrial transcript. Nucleic Acids Res. 2002;30:1895–901.View ArticlePubMedPubMed CentralGoogle Scholar
- Nutt SL, Hoo KH, Rampersad V, Deverill RM, Elliott CE, Fletcher EJ, et al. Molecular characterization of the human EAA5 (GluR7) receptor: a high-affinity kainate receptor with novel potential RNA editing sites. Receptors Channels. 1994;2:315–26.PubMedGoogle Scholar
- Klimek-Tomczak K, Mikula M, Dzwonek A, Paziewska A, Karczmarski J, Hennig E, et al. Editing of hnRNP K protein mRNA in colorectal adenocarcinoma and surrounding mucosa. Br J Cancer. 2006;94:586–92.View ArticlePubMedPubMed CentralGoogle Scholar
- Jepson JE, Reenan RA. RNA editing in regulating gene expression in the brain. Biochim Biophys Acta. 2008;1779:459–70.View ArticlePubMedGoogle Scholar
- Tonkin LA, Saccomanno L, Morse DP, Brodigan T, Krause M, Bass BL. RNA editing by ADARs is important for normal behavior in Caenorhabditis elegans. EMBO J. 2002;21:6025–35.View ArticlePubMedPubMed CentralGoogle Scholar
- Palladino MJ, Keegan LP, O‘Connell MA, Reenan RA. A-to-I pre-mRNA editing in Drosophila is primarily involved in adult nervous system function and integrity. Cell. 2000;102:437–49.View ArticlePubMedGoogle Scholar
- Zipeto MA, Jiang Q, Melese E, Jamieson CH. RNA rewriting, recoding, and rewiring in human disease. Trends Mol Med. 2015;21:549–59.View ArticlePubMedGoogle Scholar
- Maas S, Kawahara Y, Tamburro KM, Nishikura K. A-to-I RNA editing and human disease. RNA Biol. 2006;3:1–9.View ArticlePubMedPubMed CentralGoogle Scholar
- Kleinman CL, Majewski J. Comment on “Widespread RNA and DNA sequence differences in the human transcriptome”. Science. 2012;335:1302. author reply 1302.View ArticlePubMedGoogle Scholar
- Lin W, Piskol R, Tan MH, Li JB. Comment on “Widespread RNA and DNA sequence differences in the human transcriptome”. Science. 2012;335:1302. author reply 1302.View ArticlePubMedGoogle Scholar
- Pickrell JK, Gilad Y, Pritchard JK. Comment on “Widespread RNA and DNA sequence differences in the human transcriptome”. Science. 2012;335:1302. author reply 1302.View ArticlePubMedGoogle Scholar
- Schrider DR, Gout JF, Hahn MW. Very few RNA and DNA sequence differences in the human transcriptome. PLoS One. 2011;6, e25842.View ArticlePubMedPubMed CentralGoogle Scholar
- Peng Z, Cheng Y, Tan BC, Kang L, Tian Z, Zhu Y, et al. Comprehensive analysis of RNA-Seq data reveals extensive RNA editing in a human transcriptome. Nat Biotechnol. 2012;30:253–60.View ArticlePubMedGoogle Scholar
- Park E, Williams B, Wold BJ, Mortazavi A. RNA editing in the human ENCODE RNA-seq data. Genome Res. 2012;22:1626–33.View ArticlePubMedPubMed CentralGoogle Scholar
- Bahn JH, Lee JH, Li G, Greer C, Peng G, Xiao X. Accurate identification of A-to-I RNA editing in human by transcriptome sequencing. Genome Res. 2012;22:142–50.View ArticlePubMedPubMed CentralGoogle Scholar
- Ramaswami G, Lin W, Piskol R, Tan MH, Davis C, Li JB. Accurate identification of human Alu and non-Alu RNA editing sites. Nat Methods. 2012;9:579–81.View ArticlePubMedPubMed CentralGoogle Scholar
- Ramaswami G, Zhang R, Piskol R, Keegan LP, Deng P, O‘Connell MA, et al. Identifying RNA editing sites using RNA sequencing data alone. Nat Methods. 2013;10:128–32.View ArticlePubMedPubMed CentralGoogle Scholar
- Zhang Q, Xiao X. Genome sequence-independent identification of RNA editing sites. Nat Methods. 2015;12:347–50.View ArticlePubMedPubMed CentralGoogle Scholar
- Chen JY, Peng Z, Zhang R, Yang XZ, Tan BC, Fang H, et al. RNA editome in rhesus macaque shaped by purifying selection. PLoS Genet. 2014;10, e1004274.View ArticlePubMedPubMed CentralGoogle Scholar
- Danecek P, Nellaker C, McIntyre RE, Buendia-Buendia JE, Bumpstead S, Ponting CP, et al. High levels of RNA-editing site conservation amongst 15 laboratory mouse strains. Genome Biol. 2012;13:26.View ArticlePubMedGoogle Scholar
- St Laurent G, Tackett MR, Nechkin S, Shtokalo D, Antonets D, Savva YA, et al. Genome-wide analysis of A-to-I RNA editing by single-molecule sequencing in Drosophila. Nat Struct Mol Biol. 2013;20:1333–9.View ArticlePubMedGoogle Scholar
- Li Q, Wang Z, Lian J, Schiøtt M, Jin L, Zhang P, et al. Caste-specific RNA editomes in the leaf-cutting ant Acromyrmex echinatior. Nat Commun. 2014;5:4943.View ArticlePubMedPubMed CentralGoogle Scholar
- Zhao HQ, Zhang P, Gao H, He X, Dou Y, Huang AY, et al. Profiling the RNA editomes of wild-type C. elegans and ADAR mutants. Genome Res. 2015;25:66–75.View ArticlePubMedPubMed CentralGoogle Scholar
- Picardi E, Pesole G. REDItools: high-throughput RNA editing detection made easy. Bioinformatics. 2013;29:1813–4.View ArticlePubMedGoogle Scholar
- Parkhomchuk D, Borodina T, Amstislavskiy V, Banaru M, Hallen L, Krobitsch S, et al. Transcriptome analysis by strand-specific sequencing of complementary DNA. Nucleic Acids Res. 2009;37, e123.View ArticlePubMedPubMed CentralGoogle Scholar
- Li H, Durbin R. Fast and accurate short read alignment with Burrows-Wheeler transform. Bioinformatics. 2009;25:1754–60.View ArticlePubMedPubMed CentralGoogle Scholar
- Fumagalli D, Gacquer D, Rothe F, Lefort A, Libert F, Brown D, et al. Principles Governing A-to-I RNA Editing in the Breast Cancer Transcriptome. Cell Rep. 2015;13:277–89.View ArticlePubMedGoogle Scholar
- Yang XZ, Chen JY, Liu CJ, Peng J, Wee YR, Han X, et al. Selectively Constrained RNA Editing Regulation Crosstalks with piRNA Biogenesis in Primates. Mol Biol Evol. 2015;32:3143–57.PubMedPubMed CentralGoogle Scholar
- Fresard L, Leroux S, Roux PF, Klopp C, Fabre S, Esquerre D, et al. Genome-Wide Characterization of RNA Editing in Chicken Embryos Reveals Common Features among Vertebrates. PLoS One. 2015;10, e0126776.View ArticlePubMedPubMed CentralGoogle Scholar
- Langmead B, Salzberg SL. Fast gapped-read alignment with Bowtie 2. Nat Methods. 2012;9:357–9.View ArticlePubMedPubMed CentralGoogle Scholar
- Kim D, Pertea G, Trapnell C, Pimentel H, Kelley R, Salzberg SL. TopHat2: accurate alignment of transcriptomes in the presence of insertions, deletions and gene fusions. Genome Biol. 2013;14:R36.View ArticlePubMedPubMed CentralGoogle Scholar
- Wu TD, Nacu S. Fast and SNP-tolerant detection of complex variants and splicing in short reads. Bioinformatics. 2010;26:873–81.View ArticlePubMedPubMed CentralGoogle Scholar
- Kim D, Langmead B, Salzberg SL. HISAT: a fast spliced aligner with low memory requirements. Nat Methods. 2015;12:357–60.View ArticlePubMedPubMed CentralGoogle Scholar
- Li H, Handsaker B, Wysoker A, Fennell T, Ruan J, Homer N, et al. Genome Project Data Processing S: The Sequence Alignment/Map format and SAMtools. Bioinformatics. 2009;25:2078–9.View ArticlePubMedPubMed CentralGoogle Scholar
- Minoche AE, Dohm JC, Himmelbauer H. Evaluation of genomic high-throughput sequencing data generated on Illumina HiSeq and genome analyzer systems. Genome Biol. 2011;12:R112.View ArticlePubMedPubMed CentralGoogle Scholar
- van Gurp TP, McIntyre LM, Verhoeven KJ. Consistent errors in first strand cDNA due to random hexamer mispriming. PLoS One. 2013;8, e85583.View ArticlePubMedPubMed CentralGoogle Scholar
- DePristo MA, Banks E, Poplin R, Garimella KV, Maguire JR, Hartl C, et al. A framework for variation discovery and genotyping using next-generation DNA sequencing data. Nat Genet. 2011;43:491–8.View ArticlePubMedPubMed CentralGoogle Scholar
- Shoemaker JS, Painter IS, Weir BS. Bayesian statistics in genetics: a guide for the uninitiated. Trends Genet. 1999;15:354–8.View ArticlePubMedGoogle Scholar
- Kent WJ. BLAT--the BLAST-like alignment tool. Genome Res. 2002;12:656–64.View ArticlePubMedPubMed CentralGoogle Scholar
- Alon S, Eisenberg E. Identifying RNA editing sites in miRNAs by deep sequencing. Methods Mol Biol. 2013;1038:159–70.View ArticlePubMedGoogle Scholar
- Benjamini Y, Drai D, Elmer G, Kafkafi N, Golani I. Controlling the false discovery rate in behavior genetics research. Behav Brain Res. 2001;125:279–84.View ArticlePubMedGoogle Scholar
- NCBI Gene Expression Omnibus http://www.ncbi.nlm.nih.gov/geo/. Accessed 24 Jul 2016.
- Hu H, Ji L, Li C, Nygaard S; Schiøtt M, Wurm Y, Zhang G, Zhou J, Qiu F, Rasmussen M, Pan H, Hauser F, Krogh A, Grimmelikhuijzen CJ, Wang J, Boomsma JJ. Genome data from the leaf-cutting ant (Acromyrmex echinatior). GigaSci. 2011; http://dx.doi.org/10.5524/100011
- NCBI GM12878 Human Cell Line DNA-seq Reads Database ftp://ftp-trace.ncbi.nih.gov/1000genomes/ftp/technical/pilot2_high_cov_GRCh37_bams/data/NA12878/alignment/. Accessed 24 Jul 2016.
- UCSC Human Genome RNA-seq Reads Database http://hgdownload.cse.ucsc.edu/goldenpath/hg19/encodeDCC/wgEncodeCshlLongRnaSeq. Accessed 24 Jul 2016.
- RES-Scanner Project Home Page https://github.com/ZhangLabSZ/RES-Scanner. Accessed 24 Jul 2016.
- Wang Z, Lian J, Li Q, Zhang P, Zhou Y, Zhan X, Zhang G. Software and supporting material for: “RES-Scanner: a software package for genome-wide identification of RNA-editing sites”. GigaScience Database. 2016; http://dx.doi.org/100211