A comprehensive review of web-based resources of non-coding RNAs for plant science research.

Non-coding RNAs (ncRNAs) are transcribed from genome but not translated into proteins. Many ncRNAs are key regulators of plants growth and development, metabolism and stress tolerance. In order to make the web-based ncRNA resources for plant science research be more easily accessible and understandable, we made a comprehensive review for 83 web-based resources of three types, including genome databases containing ncRNA data, microRNA (miRNA) databases and long non-coding RNA (lncRNA) databases. To facilitate effective usage of these resources, we also suggested some preferred resources of miRNAs and lncRNAs for performing meaningful analysis.

Introduction

Non-coding RNAs (ncRNAs) are transcribed from genome but will not be translated into proteins 1. Some of ncRNAs play key roles in plants growth and development, metabolism and stress tolerance 2, 3. At present, the research of plant non-coding RNA mainly focuses on two areas: small RNAs with lengths of less than 50 nucleotides (nt), including microRNAs (miRNAs) and small interfering RNAs (siRNAs), and long non-coding RNAs (lncRNAs) normally with lengths of more than 200 nt, including long intergenic non-coding RNAs (lincRNAs), long intronic non-coding RNAs, circular RNAs (circRNAs), and circular intronic RNAs (ciRNAs) 4.

Up till now, more than 100 plant genomes have been sequenced, including more than 60 crop plants and several model plants 5. Most of them are provided with genome databases, such as Arabidopsis 6, 7, wheat 8, rice 9, 10, soybean 11 and so on. Most ncRNAs of crops or model plants could be found by using these databases. And genome databases are useful for exploring the functions of ncRNAs and for searching their genomic contents 12, 13. Thus genome databases are widely used as references in ncRNA studies.

MicroRNAs (miRNAs) are a class of endogenous single-strand ncRNAs sized from 18 to 25 nt, which participate in the regulation of gene expression at post transcriptional level in plants 14. Plant miRNAs generally bind to the coding regions of the genes through intensively matched complementary sites that induce cleavage or translational repression of specific protein-coding genes 15. They affect many biological processes, such as cell cycle, differentiation, and metabolism 2, 3. And the increasing evidence indicates that miRNAs play key roles in developments of organs and stress responses 16.

Long non-coding RNA (lncRNA) is a class of endogenous ncRNAs with lengths of more than 200 nt but do not have coding potential 17, 18. Though lncRNAs were ignored in early studies since their low expression and sequence conservation, emerging evidence indicates that many lncRNAs are important in diverse biological processes, such as regulation of flowering time, gene silencing, root organogenesis, seedling photomorphogenesis, and reproduction 19-22.

In the following parts of the work, we will review some web-based resources of genomes, miRNAs and lncRNAs. Then we briefly describe the analysis methods of ncRNAs, and recommend some preferred resources for the related analyses. Finally, we briefly introduce a useful web-based plant comprehensive database, the Joint Genome Institute (JGI) Genome Portal 23, 24.

Resources of plant genomes

As listed in Table 1, we will first introduce 25 resources for genomic information.

Table 1

Web-based resource for plant genomes

Database/tool	URL	Description	Ref.
TAIR	http://www.arabidopsis.org/	A resource for Arabidopsis genome information	6, 7
Wheat	http://www.wheatgenome.org/	A resource for wheat genome information	8
Rice	http://rice.plantbiology.msu.edu/	A resource for rice genome information	9, 10
MaizeGDB	https://www.maizegdb.org/	A database for Maize genetics and genomics	25, 26
SoyBase	http://soybase.org/	A database for the USDA-ARS soybean genetics and genomics	27
funRiceGenes	https://funricegenes.github.io/	A comprehensive database of functionally characterized rice genes	28
GO	http://geneontology.org/	A web for Gene Ontology resource	29, 30
KEGG	http://www.kegg.jp/	A Synthesis web for Kyoto Encyclopedia of Genes and Genomes	31
KOBAS	http://kobas.cbi.pku.edu.cn/	A wed for gene/protein functional annotation (Annotate module) and functional gene set enrichment(Enrichment module)	32, 33
RiceWiki	http://ricewiki.big.ac.cn/	A wiki-based, publicly editable and open-content platform for community curation of genes and Omics Knowledge in rice	34-37
BRAD	http://brassicadb.org/	A database for genome scale genetic and genomic data of Brassica crops	38, 39
MIPS PlantsDB	http://mips.helmholtz-muenchen.de/plant/genomes.jsp/	A database for comparative plant genome research.	40, 41
SorGSD	http://sorgsd.big.ac.cn/	A sorghum genome SNP database	42
MtGEA	http://bioinfo.noble.org/gene-atlas/	A web server for analyzing the Medicago transcriptome	43
DsTRD	http://bi.sky.zstu.edu.cn/DsTRD/home.php/	A comprehensive resource for the basic and applied for Salvia Miltiorrhiza researchers	44
SoyFN	http://nclab.hit.edu.cn/SoyFN/	A database for soybean functional gene networks and miRNA functional networks	11
GraP	http://structuralbiology.cau.edu.cn/GraP/	A database for functional annotation, gene family classifications, protein-protein interaction networks, co-expression networks and microRNA-target pairs of Gossypium raimondii	45
STPD	http://me.lzu.edu.cn/stpd/	A comprehensive database for salt-tolerant adaption and poplar genomics of tree	12
TFGD	http://ted.bti.cornell.edu/	A database provides a comprehensive resource to store, query, mine, analyze, visualize and integrate large-scale tomato functional genomics data sets	13
LegumeIP	http://plantgrn.noble.org/LegumeIP/	An integrative database for comparative genomics and transcriptomics of model legumes	46
Plant Reactome	http://plantreactome.gramene.org/	A freely accessible database of plant metabolic and regulatory pathways	47-49
Plant-PrAS	http://plant-pras.riken.jp/	A database of physicochemical and structural properties and novel functional regions in plant proteomes	50, 51
PlantGDB	http://www.plantgdb.org/	A resource for comparative plant genomics	52
CuGenDB	http://www.icugi.org	A genomic database for Cucurbit species	53
Ensembl Plants	http://plants.ensembl.org	An integrative resource for sequenced plant species (currently 38)	54

The Arabidopsis Information Resource (TAIR) 6, 7, International Wheat Genome Sequencing Consortium (IWGSC) 8, the Institute for Genomic Research (TIGR)/Michigan State University (MSU) Rice Annotation Database 9, 10, MaizeGDB 25, 26, SoyBase 27 are databases for Arabidopsis, wheat, rice, maize, and soybean genome, respectively. Users can download the genome sequences, annotation of genes and other information from these databases.

The funRiceGenes collected about 2,800 functional enrichment rice genes and about 5,000 members of different gene families 28.This database provides a more detailed data set to rice gene functions for researchers, including biological pathway, gene functional annotation.

The Gene Ontology (GO) is a cluster and correlation analysis database for describing gene products which was built in 1998. With development of sequencing study, increasingly number of databases has been incorporated in the GO Consortium (GOC), including several of green plant, animal, and microbial genomes. The GO has three structured categories that describe gene products in terms of their associated biological processes, cellular components and molecular functions 29, 30. In ncRNA study, GO could be used to annotate ncRNAs and their cellular function analysis.

Kyoto Encyclopedia of Genes and Genomes (KEGG) is a network resource database for understanding the functions of proteins and utilities of the biological system, such as cell, organism and ecosystem. One of the features of the KEGG database is that the genetic catalogue from the genome which has been sequenced and linked to higher levels of cell, species and ecosystem level 31. This database is always used to annotate targets of ncRNAs and perform metabolic pathway analysis.

KOBAS includes 1,327 species from 5 pathway databases (KEGG PATHWAY, PID, BioCyc, Reactome and Panther) and 5 human disease databases (OMIM, KEGG DISEASE, FunDO, GAD and NHGRI GWAS Catalog) 32, 33. In plant non-coding RNA study, it could be used to annotate target gene of non-coding RNA function, including GO and KEGG annotation. Furthermore, KOBAS using statistical methods to identify metabolic pathways significantly enriched 32, 33.

RiceWiki 34-37 is a community-curated resource of rice knowledge. More than 1000 manually curated genes are validated as high quality in RiceWiki. Gene Expression Profiles retrieved from the Information Commons for Rice (IC4R) 35, 37 are integrated in RiceWiki.

The Brassica Database (BRAD) is a genomic database of Brassica rapa and Arabidopsis thaliana 38. And BRAD v2.0 39 has integrated 11 more Brassicaceae genomes, including Arabidopsis lyrata, Aethionema arabicum, Brassica oleracea, Brassica napus and so on. This database provides some conserved genomic information about Brassicaceae species, from the molecular level to the genomic level. And a module of Generic Synteny Browser (GBrowse_syn) provides search related functions for seeking syntenic and non-syntenic orthologs and their annotation and sequences 38, 39.

Munich Information Center for Protein Sequences PlantsDB (MIPS PlantsDB) focuses on the bioinformatics analysis methods of plant genomes 40, 41. It aims to provide some related resource for each plant species. Meanwhile this database provides a search platform for plant genome-related research. Plant Genome and Systems Biology PlantsDB (PGSB PlantsDB) is a database for framework to facilitate comparative plant genome research. It provides services of keyword search options and BLAST sequence search function. It mainly includes complex Triticum aestivum L. genome data, especially for wheat, rye, and barley 41.

SorGSD is a database for sorghum genome SNP 42. SorGSD contains 62 million SNPs information from 48 groups of different sorghum varieties. Using web interface, it provides an easy way to efficiently browse, search, and analyze SNP. SorGSD allows users to retrieve the SNPs information and their relevant annotations for each sorghum genome SNP.

The Medicago truncatula Gene Expression Atlas (MtGEA) aims to analyze the Medicago transcriptome, and it also offers search and analysis tools on the web platform for Medicago transcriptome 43.

Danshen Transcriptional Resource Database (DsTRD) provides information on transcript sequences and functional annotations including protein-coding RNAs, lncRNAs, other ncRNA, miRNAs and phasiRNAs 44.

SoyFN is a resource about soybean miRNA functional networks and functional gene networks 11. It also can be used to seek, download, and analyze the functional networks of soybean miRNAs and functional genes, including their KEGG pathways, GO annotations and 3'-UTR sequences. In addition, it provides some useful tools, such as including SoySearch, promoter motif scanning, Genome Browser, eFP Browser and ID mapping 11.

Platform of Functional Genomics Analysis in Gossypium raimondii (GraP) is a genome database for Gossypium raimondii which is one of the most importance natural fiber and oil crops 45. This database is constructed to provide visualization tools, integration and multi-dimensional analysis which provide service of gene family classifications, functional annotation, microRNA-target pairs, co-expression networks and protein-protein interaction networks 45.

The Salinity Tolerant Poplar Database (STPD) contains data of genomic sequences, genes and their functional information, single nucleotide polymorphisms information, transposable elements, simple sequence repeats and ncRNA sequences from Populus euphratica, gene expression data between Populus euphratica and Populus tomentosa, and whole-genome alignments between Populus trichocarpa, Populus euphratica and Salix suchowensis 12. And this database provides data mining and useful searching tools, such as genome browser, BLAST servers and genome alignments viewer and so on. These tools can be used to identify similar sequences, visualize genome alignments, and browse genome regions 12.

Tomato Functional Genomics Database (TFGD), a tomato functional genomics database, which collects RNA-seq data, microarrays, small RNA data, metabolite of tomatoes and can be used to store, retrieve, analyze, visualize and integrate large-scale tomato functional annotation genomics data sets, as well as to download all the analysis results 13.

LegumeIP 46 is for research genome evolution and gene functions in legumes. The database includes three model legumes: Medicago truncatula, Glycine max, Lotus japonicus and two reference plant species, Arabidopsis thaliana and Populus trichocarpa. LegumeIP also provides visualization and comprehensive search tools that enable flexible queries based on gene annotation, relative gene expression, and gene family 46.

Plant Reactome is a free open-source curated plant pathway database, including 1,025 reactions associated with 1,173 proteins, 222 rice pathway terms, 907 small molecules 47. The database provides bioinformatics tools for the visualization, analysis and interpretation of pathway knowledge to support genome analysis, genome annotation and basic research 47-49.

The Plant Protein Annotation Suite database (Plant-PrAS) includes 6 typical plant species, Arabidopsis thaliana, Glycine max (soybean), Populus trichocarpa (poplar), Oryza sativa (rice), Physcomitrella patens (moss) and Cyanidioschyzon merolae (alga) 50, 51. Plant-PrAS provides multiple physicochemical and secondary structural parameters of protein sequences about these species 50, 51.

PlantGDB is a genome database encompassing sequence data for green plants (Viridiplantae). It provides transcript assemblies annotated for more than 100 plant species 52.

Cucurbit Genomics Database (CuGenDB) is a database for several important crops of Cucurbit species, including cucumber, watermelon, melon and so on 53. This database not only provides genomic sequences, EST and RNA-seq of Cucurbit species, but also provides tools of searching gene information, BLAST, JBrowse, Pathway and GO enrichment and so on, as well as functions for downloading genome and annotation, genome resequencing, marker, EST, unigene and pathway 53.

Ensembl Plants is an integrative resource for sequenced plant species (with current version 38) 54. This database provides information of genome sequence, functional annotation, gene models and polymorphic loci. In this database, various kinds of additional information are provided which including individual genotypes, population structure, phenotype data and linkage. And it also provides comparative analyses of whole genome and protein sequences; and service for building genome alignments and gene trees. This database is updated 4-5 times a year as well as added release genomes and information of new species 54.

Resources of miRNAs

We will next introduce 34 resources of miRNAs listed in Table 2.

Table 2

Web-based resources for miRNAs

Database/tool	URL	Description	Ref.
miRBase	http://www.mirbase.org/	The official database of miRNAs in different species	55-58
Rfam	http://rfam.xfam.org/	Non-coding RNA genes, structured cis-regulatory elements and self-splicing RNAs	59
miRTarBase	http://miRTarBase.mbc.nctu.edu.tw/	A database for experimentally validated microRNA-target	60-63
mirPub	http://www.microrna.gr/mirpub/	A database for searching microRNA publications	64
PatScan	http://www.mcs.anl.gov/home/overbeek/PatScan/HTML/pats~dn.html/	A Web-based system enablesusers to search for nucleotide sequences for secondary structure (hairpins, pseudoknots, hammerheads and so on) and protein sequences for known motifs	65
PsRobot	http://omicslab.genetics.ac.cn/psRobot/	A web-based easy-to-use tool dedicated to the identification of smRNAs with stem-loop shaped precursors (such as microRNAs and short hairpin RNAs) and their target genes/transcripts	66
psRNATarget	http://plantgrn.noble.org/psRNATarget/	A plant small RNA target analysis server	67
CleaveLand	http://www.bio.psu.edu/people/faculty/Axtell/AxtellLab/Software.htm/	A pipeline for using degradome data to find cleaved small RNA targets	68
miRNApath	http://lgmb.fmrp.usp.br/mirnapath/	A database of miRNAs, target genes and metabolic pathways	69
TarBase	http://microrna.gr/tarbase/	A database for microRNA targets in several animal species of central scientific interest, plants and viruses	70, 71
PeTMbase	http://petmbase.org/	A Database of Plant Endogenous Target Mimics (eTMs)	72
AtmiRNET	http://AtmiRNET.itps.ncku.edu.tw/	A resource for reconstructing regulatory networks of Arabidopsis miRNAs	73
PmiRExAt	http://pmirexat.nabi.res.in/	A database and web applications for plant miRNA expression atlas	74
tasiRNAdb	http://bioinfo.jit.edu.cn/tasiRNADatabase/	A resource for the sequences of ta-siRNA regulatory pathway-related microRNAs, TASs, ta-siRNAs and ta-siRNA targets, and for the cascading relations among them	75
comTAR	http://rnabiology.ibr-conicet.gov.ar/comtar/	A web tool for the prediction and characterization of conserved miRNA targets in plants	76
miR-PREFeR	https://github.com/hangelwen/miRPREFeR/	A web tool for plant miRNA prediction by using small RNA-Seq data	77
ASRP	http://asrp.cgrb.oregonstate.edu/	A database for microRNAs and short-interfering RNAs (siRNAs) of Arabidopsis	78
PMRD	http://bioinformatics.cau.edu.cn/PMRD/	A database of plant miRNA data deposited in public databases, gleaned from the recent literature, and data generated in-house	79
starBase	http://starbase.sysu.edu.cn/	A database for developing to facilitate the comprehensive exploration of miRNA-target interaction maps from CLIP-Seq and Degradome-Seq data	80
miRNEST	http://mirnest.amu.edu.pl/	A database for integrative approach in microRNA search and annotation	81, 82
siRNAdb	http://siRNA.cgb.ki.se/	A database of siRNA sequences	84
pssRNAMiner	http://bioinfo3.noble.org/pssRNAMiner/.	A plant short small RNA regulatory cascade analysis server	85
TAPIR	http://bioinformatics.psb.ugent.be/webtools/tapir/	A web server for the prediction of plant microRNA targets	86
RNA-hybrid	http://bibiserv.techfak.uni-bielefeld.de/rnahybrid/	A program for predicts multiple potential binding sites of miRNAs in large target RNAs	87
PMTED	http://pmted.agrinome.org/	A database for retrieving and analysis expression profiles of miRNA targets represented	88
PASmiR	http://pcsb.ahau.edu.cn:8080/PASmiR/	A database for miRNA molecular regulation in plant response to abiotic stress	89
AHD	http://ahd.cbi.pku.edu.cn/help.php	A database of collection of hormone related genes of the model organism Arabidopsis thaliana	90, 91
miRFANs	http://www.cassavagenome.cn/mirfans/	An integrated database for Arabidopsis thaliana microRNA function annotations	92
MPSS	http://mpss.udel.edu/	A resource for mRNA and small RNA of Arabidopsis, rice, grape and Magnaporthe grisea and the rice blast fungus	93
plant_snoRNA DB	http://www.scri.sari. ac.uk/plant_snoRNA/	A database for information of small nucleolar RNAs from Arabidopsis and 18 other plant species	94
miSolRNA	http://www.misolrna.org/	A database for defining physiological modes of action of regulatory process underlying the metabolism of the tomato fruit	95
PmiRKB	http://bis.zju.edu.cn/pmirkb/	A database including four major functional modules: SNP, Pri-miR, MiR-Tar and Self-reg module.	96
PlantDARIO	http://plantdario.bioinf.uni-leipzig.de/	A web server for qualitative analysis of small RNA-seq data in plants	98
OmiRas	http://tools.genxpro.net/omiras/	A Web server for the annotation, comparison and visualization of interaction networks of differential expression non-coding RNAs	99

The miRBase reports the miRNAs in all species officially, including 4,886 plant species 55, 56. Up to now, the miRBase has been updated to release 21 57.We can obtain the sequences of pre-miRNAs and mature miRNAs, the secondary structures of pre-miRNAs, and related literature from the miRBase 58.

Rfam is a database of RNA families which collected a multiple sequence alignment, a covariance model (CM) and a consensus secondary structure. Rfam provides three broad functional classes including structured cis-regulatory elements, non-coding RNA genes and self-splicing RNAs. Users can use it to identify ncRNA family, including families of miRNAs 59.

The miRTarBase has accumulated more than 360,000 miRNA-target interactions (MTIs) 60. By providing experimentally validated interaction for miRNA-target, miRTarBase aims to elucidate the role of miRNAs in different conditions and in different species including Arabidopsis thaliana, Oryza sativa 60-63.

The mirPub 64 can be used for searching microRNA publications. Because man-made causes, miRNA always named in the literature of deviation from the standard naming convention, even the official nomenclature is also developing. This database provides powerful search service to help users search relevant miRNA literature 64.

PatScan is one of the earliest programs used for searching complementary target genes 65. PatScan can also be used to search the complementary matching sequences that allow GU wobble, mismatch and bulged nucleotide (insertion/deletion) between miRNAs and target sites 65.

PsRobot is a web-based plant small RNA meta-analysis toolbox which can be used to solve problems partially in plant miRNA and target prediction 66.

The psRNATarget is a server for analyzing plant small RNA target, containing recent study in plant miRNA target recognition 67. The psRNATarget server is an efficient distributed computing back-end pipeline that runs on a Linux cluster which is designed for high-throughput analysis of next-generation data 67.

CleaveLand is designed to detect cleaved miRNA targets from degradome data. CleaveLand takes small RNAs, degradome sequences and an mRNA database as input and outputs small RNA targets 68.

The miRNApath is an online database for linking miRNAs to metabolic pathways using miRNA: target relations 69. Users can enter information to search for relevant target genes and metabolic pathways 69.

In the TarBase 70 and TarBase 5.0 71, the results are associated with pathway and GO terms, together with regulating miRNAs predicted by different computational methods and the target site information 70, 71.

Plant Endogenous Target Mimicry (PeTMbase) aims to provide a method for endogenous miRNA target mimics (eTM) in plants. It is a detectable database with 2,728 eTM sequences predicted computationally from eleven different green plant species 72.

AtmiRNET is a database of Arabidopsis miRNAs with emphasis on miRNA target recognition, promoter identification, functional enrichment of target genes, and detection of cis- and trans-elements 73. It also supports the analysis of elucidating of functional enrichment analysis and correlations between the expression levels of miRNAs and their targets 73.

Plant miRNA Expression Atlas (PmiRExAt) is a web-based resource comprises of miRNA expression profile and searching tool for 1,859 wheat, 2,330 rice and 283 maize miRNAs 74. The database interface offers open and easy access to miRNA expression profiles and helps to identify miRNAs expressing 74.

TasiRNAdb is a database for storing the sequences of ta-siRNA regulatory pathway-related microRNAs, ta-siRNA, TASs and ta-siRNA targets 75. This database included 583 pathways which are regulated by ta-siRNA from 18 species. TasExpAnalys in this database is used for mapping submitted small RNAs and degradome libraries to input or TasiRNAdb's TAS sequences and performing sRNA phasing analysis and TAS cleavage analysis 75.

Conserved microRNA targets (ComTAR) is a network resource tool for the prediction of miRNA targets 76. This database includes transcript data of 33 angiosperms which are designed to create data sets of potential miRNA targets. This database includes describing miRNA target information, functions and evolutionary conservation 76.

The miRNA PREdiction From small RNA-Seq data (MiR-PREFeR) is a plant microRNA prediction tool 77. It can analyze miRNAs from small RNA-Seq data by utilizing expression modes of miRNAs. It follows the criteria for plant microRNA annotation to precisely predict plant miRNAs from small RNA-Seq data samples of the same species 77.

Arabidopsis Small RNA Project (ASRP) is a search tool of Arabidopsis small RNA which can describe and analyze the major classes of endogenous small RNAs in Arabidopsis. It offers sequences of small RNAs cloned from multiplicate Arabidopsis genotypes, including 1,920 unique sequences of miRNA and siRNA 78.

The Plant miRNA Database (PMRD) is a plant miRNA database which including 8,433 miRNAs assembled from 121 plant species and integrate plant miRNA data from public databases, recent researches 79. This database contains small RNA sequence information, target sites, secondary structure, a genome browser and expression profiles 79.

StarBase is a database which collects extensive researches of miRNA-target interaction maps from Degradome-Seq and CLIP-Seq data, including high-throughput sequencing data originated from 10 degradome-Seq and 21 CLIP-Seq experiments from 6 organisms 80. More than 2 million cleaved target clusters of plants and more than 1 Ago-binding clusters of animal are identified. From these clusters, 400,000 and 66,000 miRNA-target regulatory relationships from Degradome-Seq and CLIP-Seq data respectively are identified. This database also can be used to seek novel miRNA target sites from Degradome-Seq and CLIP-Seq data and to annotate their targets by GO terms and pathways 80.

MiRNEST is an integrative miRNAs resource which integrates external miRNA data from 13 databases and related researches, which includes small RNA sequencing data, miRNA sequences, polymorphisms, expression and targets data as well as associates to external miRNA resources from 544 species, including animals, plants and viruses 81. The miRNA predictions performed on Expressed Sequence Tags of 202 plant and 225 animal species are the core of this database 81. MiRNEST 2.0 added new miRNA predictions program from deep sequencing libraries 82, results of pre-miRNA classification with HuntMi 83, data from the analyses of plant degradome, and miRNA splice sites information and functions of download/upload options. Furthermore MiRNEST 2.0 improved the user interface to make it more convenient to browse through miRNA records 81, 82.

The siRNA Database (SiRNAdb) provides a gene-centric comment of siRNA experimental data, including siRNAs predicted to be of high efficacy and siRNAs of known efficacy by a combination of methods 84. The database is developed for evaluating a siRNA's potential for non-specific and inhibition effects 84.

PssRNAMiner is designed to identify both the potential phase-initiator and the clusters of phased small RNAs 85. To identify phase-initiators, the user should submit no less than one small RNA sequence as candidate phase initiator. The pssRNAMiner can be used to identify phased small RNA clusters as ta-siRNA candidates by evaluating the P-values of hypergeometric distribution 85.

TAPIR is performed for the prediction of plant microRNA targets 86. The server provides a fast and precise algorithm to retrieve plant miRNA target sites. The precise option guarantees to find less perfectly paired miRNA-target duplexes though it is much slower 86.

RNA-hybrid is used for finding the minimum free energy hybridization of ncRNA 87. The tool is primarily considered as a means for microRNA target prediction 87.

Plant MiRNA Target Expression Database (PMTED) is a database for retrieving and analyzing the expression of miRNA target genes based on the miRNA data in existing literatures 88. Function for miRNAs and their target sequences, and differential expression profiles can be easily obtained by using this database. And it can be used to predict conserved and novel miRNAs and their targets with their corresponding expression profile retrieval. It also provides service of browsing functions in a global Meta-network among bioprocesses, species, conditions, and miRNAs, meta-terms collected from well annotated microarray experiments, which are displayed by a Cytoscape Web-based graphical interface 88.

PASmiR is a database for miRNA molecular regulation network in different plant under abiotic stresses 89. It contains more than two hundred published studies, representing more than one thousand regulatory relationships between 35 abiotic stresses and 682 miRNAs in 33 plant species. Users can search the miRNA-stress regulatory entries with keywords of plant species, abiotic stress, and miRNA identifier. In this database, users are provided with detailed regulation information of a specific miRNA, including miRNA expression pattern, the name of species, stress name, miRNA identifier, detection method for miRNA expression, a reference literature, and target gene(s) of the miRNA extracted from the corresponding reference or the miRBase 89.

Arabidopsis Hormone Database (AHD) provides a comprehensive and systematic view of genes participating in plant hormonal regulation, and describes morphological phenotypes controlled by plant hormones 90. AHD 2.0 adds predicted miRNA complementary site(s) and integrated genes that genetically interact with each Arabidopsis thaliana hormone related genes by literatures mining. This database is useful for studying miRNA mediated hormone regulation in plants 90, 91.

MiRFANs is a database for Arabidopsis thaliana miRNA function annotations, including transcription factor (TF), miRNA-target interactions and their target sites, expression profiles, pathways and genomic annotations 92. It also can be used to predict miRNA target by its various types of tools. Each miRNA target is related to pathway and GO terms and is associated with regulating miRNAs predicted information 92.

Massively Parallel Signature Sequencing (MPSS) is a signature-based transcriptional resource for mRNA and small RNA of four species (Arabidopsis, rice, grape and Magnaporthe grisea) 93. It can measure the expression levels of most genes under defined conditions and it provides information about potentially novel transcripts 93.

Plant snoRNA DB provides information on small nucleolar RNAs from Arabidopsis and 18 other plant species, including their polycistronic, intronic and the number of gene variants and many snoRNA genes from alignments of orthologous genes and gene variants from different plant species 94.

MiSolRNA collects data of tomato for integrating genetic map sites of miRNA-targeted genes, their relations with quantitative fruit metabolic loci and their expression profiles and yield associated traits 95. And it provides a metadata source to facilitate the construction of hypothesis aimed at defining physiological modes of action of regulatory process underlying the metabolism of the tomato fruit 95.

Plant miRNA knowledge base (PmiRKB) is a database mainly for miRNA information of Arabidopsis (Arabidopsis thaliana) and rice (Oryza sativa), including single nucleotide polymorphism (SNP) data to inspect the SNPs within pre-miRNAs (precursor microRNAs) and miRNA—target RNA duplexes, investigating the tissue-specific, transcriptional contexts of pre- and pri-miRNAs (primary microRNAs) and validating miRNA and target pairs 96.

PlantDARIO is a database for processing and analyzing of plant-specific small non-coding RNAs (sncRNAs) 97, 98. PlantDARIO has diverse functions, such as small RNA-seq quality control, analyses of mapping files, expression analyses of annotated sncRNAs, as well as the prediction of novel miRNAs and snoRNAs from expression analysis of user-defined loci and unknown expressed loci 98.

OmiRas is an online server to analyze differential expression of miRNAs derived from small RNA-Seq data 99. OmiRas has two main functions. The first function is static annotation, including length distribution, alignments mapping statistics, and quantification tables for each library and analysis on different expression of ncRNAs between treatment and conditions. And the second function is an interactive network of miRNAs and their target genes that are analyzed by the combination of several miRNA-mRNA interaction databases 99.

Resources of lncRNAs

In this section, we introduce 24 resources for lncRNAs, as listed in Table 3, and a new database for plant circRNAs.

Table 3

Web-based resources for lncRNAs

Database/tool	URL	Description	Ref.
lncRNAdb	http://www.lncrnadb.org/	A database for regulation mRNA and lncRNAs that have or associate with biological functions in eukaryotes	100, 101
NONCODE	http://www.noncode.org/	A database of expression and biological functions of lncRNAs	102-106
CPC	http://cpc2.cbi.pku.edu.cn/	Calculate protein-coding potential of lncRNAs and other RNAs	107, 108
NPInter	http://www.bioinfo.org/NPInter/	Functional interactions between ncRNAs and biomolecules	109, 110
CNCI	http://www.bioinfo.org/software/cnci/	A tool for classify protein-coding or non-coding transcripts	111
COME	https://github.com/lulab/COME	A tool for identification and characterization of coding potential lncRNAs	112
CPAT	http://code.google.com/p/cpat/	A tool for extracting coding and non-coding transcripts from raw transcriptome sequencing data by using logistic regression model	113
Pfamscan	https://www.ebi.ac.uk/Tools/pfa/pfamscan/	A tool for protein functional analysis	114
phyloCSF	https://github.com/mlin/PhyloCSF/wiki/	A tool for distinguish protein coding and non-coding regions	115
PLNlncRbase	http://bioinformatics.ahau.edu.cn/PLNlncRbase/	A database for lncRNAs that collected 1,187 plant lncRNAs in 43 plant species	116
PLncDB	http://chualab.rockefeller.edu/gbrowse2/homepage.html	A database of plant long non-coding	117
PNRD	http://structuralbiology.cau.edu.cn/PNRD/	A comprehensive integrated web resource for ncRNA searching, browsing, predicting, visualizing and downloading	118
LncReg	http://bioinformatics.ustc.edu.cn/lncreg/	A database developed by collecting 1,081 validated lncRNA-associated regulatory entries	119
Mfoldweb server	http://www.bioinfo.rpi.edu/applications/mfold/	A web server for nucleic acid folding and hybridization prediction	120
RNAsoft	http://www.RNAsoft.ca/	A suite of RNA secondary structure prediction and design software tools	121
RNA shapes	http://bibiserv.cebitec.uni-bielefeld.de/rnashapesstudio/	A tool for RNA structure analysis	122
RNA Movies 2	http://bibiserv.techfak.uni-bielefeld.de/rnamovies/	A powerful visualization tool for RNA secondary structure analysis	123
RNAz	http://www.rna.tbi.univie.ac.at/RNAz/	A web for RNA secondary structure prediction	123, 124
Infernal	http://infernal.janelia.org/	A tool for search and align homologous RNAs	126
PlantcircBase	http://ibi.zju.edu.cn/plantcircbase/	A database of plant circRNAs which collect reported and unpublished circRNAs of five model plants and provides services of predicting circRNAs	127
PlantCircNet	http://bis.zju.edu.cn/plantcircnet/index.php	A visualized web-based database of plant circRNA, miRNA and mRNA regulatory networks in eight model plants	128
PlantNATsDB	http://bis.zju.edu.cn/pnatdb/	A web database which includes high-throughput small RNA sequencing data investigate the biological function of natural antisense transcripts (NATs)	129.
RNAcentral	http://rnacentral.org	A database for integrating ncRNA from most organisms and providing text, sequence similarity and genome browsing functionality searching service	130, 131
ncRNAdb	http://biobases.ibch.poznan.pl/ncRNA/	A database for collecting ncRNA except microRNAs, snoRNAs and housekeeping transcripts in animals, plants, eukaryotes, eubacteria and archaea.	132, 133

The lncRNAdb is a database of lncRNAs which have or suspected to have biological functions in eukaryotes 100. There are many referenced information of RNA in each entry, including sequences, structural information, genomic context, expression, subcellular localization, conservation, functional evidence and other relevant information. The lncRNAdb can be searched by keywords, including published RNA names and aliases, sequences, species and associated protein-coding genes and so on. The lncRNAdb v2.0 101 provides reference database of more than 280 eukaryotic lncRNAs from scientific literature and add new characteristics in lncRNAdb contain a BLAST search tool and easy export of content via direct download or a REST API, the integration of Illumina Body Atlas expression profiles, nucleotide sequence information 100, 101.

NONCODE is an integrated knowledge database dedicated to non-coding RNAs, including non-coding RNAs of plants 102. This database provides the detailed information of lncRNAs, which strand, exon number, length and sequence. In addition, it also provides advanced information, which include the expression profiles, exosome expression profiles, conservation information, predicted functions and disease relations 102-106.

Coding Potential Calculator (CPC) is an available mobile-friendly web server which can be used to assess the potential protein-coding transcripts by using sequence features and provide vector machine 107. CPC2 added the function of downloadable standalone package and is much faster and more accurate 107, 108.

NPInter is a database for functional interactions between non-coding RNAs (except tRNAs and rRNAs) and biomolecules (proteins, RNAs and DNAs) 109. It also provides a convenient search tool option which is allow users seek the interactions, related publications and other information 109, 110.

Coding-Non-Coding Index (CNCI) is a tool for assortment with protein-coding or non-coding transcripts. It can use incomplete transcripts and antisense transcripts to effectively predict non-coding RNA 111.

Coding potential calculator based on multiple evidences (COME) is an accurate and stable tool for identification and characterization of coding potential lncRNAs. And this tool integrates multiple sequence-derived and experiment-based features using a method of decomposition and combination. COME can greatly improve the consistency of predication results from other coding potential calculators. In addition, COME can annotates and characterizes each predicted lncRNA transcripts with multiple lines of supporting evidence, rather than other tools provided. 112.

Coding-Potential Assessment Tool (CPAT) is a high accuracy and fast tool for extracting coding and non-coding transcripts from raw transcriptome sequencing data by using logistic regression model which base on four sequence features: open reading frame size, open reading frame coverage, Fickett TESTCODE statistic and hexamer usage bias. It also can be used online and allows users to submit FASTA or BED formatted data files of sequences 113.

Pfamscan is a tool for protein functional analysis. The possible protein sequences in each transcript will be searched homologous to the database to eliminate the possibilities of being coding genes for putative lncRNAs 114.

PhyloCSF is a tool for differentiate protein coding and non-coding regions. It can help distinguish between protein-coding and non-coding RNA in transcripts obtained by sequencing high-throughput transcriptomes 115.

PLNlncRbase is a database which provides detailed information for identified lncRNA and includes 1,187 plant lncRNAs of more than 40 species from nearly 200 published literatures 116. Users can seek plant lncRNAs by keywords of plant species or lncRNA identifier. The details of lncRNAs include species name, description of the potential biological function, lncRNA sequences and classification, expression patterns of lncRNAs, the tissue, developmental stage and condition for lncRNA expression, detection method for lncRNA expression, a reference literature, and the potential target genes of the lncRNA taken out from the original reference 116.

Plant Long non-coding RNA Database (PLncDB) includes a comprehensive genomic view of Arabidopsis lncRNAs for the plant research community 117.

Plant ncRNA Database (PNRD) contains 25,739 non-coding RNA records, including lncRNAs, tRNA, rRNA, tasiRNA, snRNA and snoRNA, etc 118. It also provides many functional search and analysis tools, including search of ncRNA keyword, literature-based function and miRNA-targets 118.

LncReg is a database for regulatory relationships of the lncRNAs, including 1,081 reported lncRNA-associated regulatory entries, 258 non-redundant lncRNAs and 571 non-redundant genes 119. It can be used to seek regulatory networks of lncRNAs and comprehensive data for bioinformatics studies and beneficial to understand the function of lncRNAs 119.

Mfold web server is a tool to predict the secondary structure of single stranded nucleic acids 120. It provides service for RNA and DNA folding and hybridization software in nucleic acids 120.

RNAsoft is a tool for prediction of the secondary structure of a pair of DNA and RNA molecules, removing unwanted secondary structure and drawing RNA strands that fold to a given input secondary structure 121. It can be used to analysis of secondary structure of lncRNA.

RNA Shapes allows users to extract several relevant structures from the folding space of RNA sequence with a single structure of minimal free energy 122.

RNA Movies 2, which enables to browse sequential paths through RNA secondary structure landscapes, can be used to visualize structural rearrangement processes of RNA 123.

RNAz is a web database for RNA secondary structure prediction 124, 125. Infernal is a tool for searching and aligning homologous RNAs 126.

PlantcircBase is a database of plant circRNAs which collect reported and unpublished circRNAs of Oryza sativa, Arabidopsis thaliana, Zea mays, Solanum lycopersicum, and Hordeum vulgare 127. And this database also provides services of predicting circRNAs from high throughput sequencing profiles input by users, for visualizing the structures of circRNAs, and analysis of functions and biogenesis mechanisms of circRNAs 127.

PlantCircNet is a visualized web-based database of plant circRNA, miRNA and mRNA regulatory networks, which can show querying detailed information of specific plant circRNAs in a user-friendly interface 128. It contain identified circRNAs in eight model plants and can find significantly overrepresented Gene Ontology categories of miRNA targets by its enrichment analysis tool. It also provides investigation of genomic annotations, sequences and isoforms of circRNAs 128.

Plant Natural Antisense Transcripts Database (PlantNATsDB) is a database for plant natural antisense transcripts (NATs) which are regulated physiological and pathological processes 129. In order to investigate the biological function of NATs, 2 million NAT pairs in 69 plant species and GO annotation and high-throughput small RNA sequencing data are integrated in this database. Small RNA can get GO annotation and their NATs function analysis in this database and their NATs regulation network could also be generated 129.

RNAcentral is a database which collects and integrates data and information of ncRNA from international consortium of established RNA sequence databases 130. The initial release contains more than 8.1 million sequences, including all major representative functional group 130. New release collects data from specialized ncRNA resources and provides a single entry point for all types of ncRNAs' sequences from most organisms, and created new species-specific identifiers of unique RNA sequences of single species. It also provides searching service of text and sequence similarity as well as genome browsing functionality 131.

Noncoding RNAs database (ncRNAdb) collected currently available sequence data on ncRNAs which have regulatory functions in prokaryotic, animal and plant cells, including regulation of chromatin structure, translational gene expression, modulation of protein function and subcellular distribution of RNAs and proteins 132, 133. New release added ncRNA sequences from eukaryotes, eubacteria and archaea. This database does not collect microRNAs, snoRNAs and housekeeping transcripts 132, 133.

A usage guide of web-based ncRNA resources

Up to now, many online resources for ncRNAs have been reported. To make it easier to use these resources, we will introduce a number of related databases and analysis tools in the general bioinformatics analysis processes of miRNAs and lncRNAs, and suggest some preferred resources.

As shown in Figure 1a, is the general process of miRNAs analysis. The first step is to obtain the required data from sequencing companies or public database such as NCBI GEO or SRA databases, and then filter the raw data to get high quality reads. There are some tools for filtering low quality reads and removing adapters in raw reads, such as Fastx-toolkit (http://hannonlab.cshl.edu/fastx_toolkit/), Trimmomatic 134, cutadapt 135. Fastx-toolkit is developed for short-reads FASTA/FASTQ files preprocessing, primarily including identical reads collapsing, read length trimming, adapter removing and so on. Fastx-toolkit is a preferred tool to filter low quantity reads and remove adapters. The second step is to align reads to mRNA databases to remove sequences from protein-coding genes, known miRNAs, other ncRNAs and repeats. Several databases such as miRBase 56, RepBase 136, and TIGR Plant Repeat Database 137 and Rfam 59 are recommended in this step. The cDNAs of the species under investigation should also be used to remove sRNAs originated from coding genes in this step. After that, in the third step, the remaining reads were aligned to the genome database. Since there could be many reads in the sequencing libraries, Bowtie 138 is suggested in this step. In the fourth step, the secondary structures of the flanking sequences of the matched loci should be predicted. Two RNA secondary structure prediction tools, Mfold 120 and RNAfold 139 are preferred in this step. Next, in the fifth step, the sequenced sRNAs from the hair-pin structures were aligned to the miRBase to examine whether the identified miRNAs were conserved or novel with BLASTN. In the sixth step, to investigate the putative functions of conserved and novel miRNAs, their target genes could be predicted using target prediction tools, such as psRobot 66, psRNATarget 67, CleaveLand 68, SeqTar 140. SeqTar 140 and PsRNATarget 67 are preferred for predicting plant miRNA target genes with and without degradome profiles, respectively. Finally, the seventh step is to predict potential functions of miRNAs. There are several tools for functional annotations, such as starBase 80, miRTarBase 60, and KOBAS 32. KOBAS is a preferred for functional annotation 32, 33 since it includes GO enrichment, and KEGG pathway, and provides a temporary storage space for user analysis. The miRTarBase is a preferred tool to search experimentally validated miRNA targets by using miRNA, target gene, pathway, validation method, disease and literature. Normally, the putative targets of miRNAs are used in functional annotation of miRNAs because miRNAs realize their functions through their regulations on target mRNAs.

Figure 1

The general bioinformatics analysis processes of miRNAs and lncRNAs, and their corresponding computational resources. The arrows indicate the suggested steps when performing miRNAs and lncRNAs bioinformatics analysis. The corresponding computational resources for the analysis steps are shown in the right panels within the dashed-line boxes. Part a and b are the general analysis processes of miRNAs and lncRNAs, respectively. The resources in bold are recommended in the corresponding analysis steps.

Figure 1b is the general analysis process of lncRNAs. Similar to miRNA analysis workflow, the first step is to filter low-quality reads from raw data. Fastx-toolkit is a preferred tool in this step. The second step is to construct spliced transcript assembly from filtered high quality reads. TopHat 141, Bowtie 138, 142, STAR 143 are preferred tools for transcript assembling with reference gene annotation and another tool, Trinity 144, is suggested for assembly without reference. In the third step, the assembled transcripts are aligned to known mRNAs to remove transcripts from known coding genes by Bowtie 2 or BWA tools. In the fourth step, the secondary structures of the remaining transcripts were predicted. RNAfold 139 and Mfold 120 are recommended for secondary structure prediction. The fifth step is to use coding potential assessment tools, such as CPC 107, CNCI 111, COME 112, CPAT 113, Pfamscan 114, phyloCSF 115, to further predict coding potentials of putative lncRNAs. CPC2 is preferred tools to predict coding potentials of lncRNAs, since CPC2 is available mobile-friendly web server and as a downloadable standalone package 107, 108. The sixth step is to examine coding potential for filtered lncRNAs by mapping them to different databases. The NCBI nt and nr database is the preferred mapping database in this step since they include sequences for all kinds of species. The seventh step is to predict potential functions of lncRNAs. NPInter 109, PNRD 118, LncRNAdb 100, PLNlncRbase 116, PLncDB 117 could be used in this step. PLNlncRbase is preferred because it is an easy-to-use database which provides detailed information for experimentally identified plant lncRNAs.

Although we introduced some related databases and tools in the analysis miRNAs and lncRNAs as shown in Figure 1, researchers can choose some other tools based on their own needs.

Genome browser

The Department of Energy (DOE) Joint Genome Institute (JGI) Genome Portal 23, 24 (http://genome.jgi.doe.gov/) is a comprehensive website that stores the high-throughput sequencing information for diverse species, including green plant, microbes, fungi, and some biological information analysis tools. JGI Phytozome (https://phytozome.jgi.doe.gov/pz/portal.html) is its sub website which is a comparative database for genomes of green plant and gene family data and analysis. There are three functional modules where users can choose and set, i.e., Species, Tools, and Download. In the module of Species, the website provides the genome of most plant species, e.g., Arabidopsis thaliana TAIR10 (Thale cress), Oryza sativa (Rice), Medicago truncatula (Barrel medic), Triticum aestivum (Common wheat), Sorghum bicolor (Cereal grass) and so on. The Tools module includes Keyword search, BLAST, BLAT, JBrowse, PhytoMine, and BioMart. Users can download a different version of plant genome sequencing information through the Download module. JGI Phytozome is updated regularly. All genomes in JGI Phytozome have been annotated with KOG, KEGG, ENZYME, Pathway, InterPro 145. In JBrowse option, users can browse the details of listed genomes, including gene site, length, and annotation.

Conclusions

NcRNAs play important regulatory roles in plants, affecting plant growth and development, disease resistances, stress tolerances, biosynthesis of secondary metabolites and many other aspects 2. With the development of modern high-throughput sequencing technology, more and more ncRNA has been reported. We summarize the web-based resources of ncRNAs and provide some preferred tools for different analysis tasks.