Keywords

1 Introduction

In the current era of genomics and next-generation sequencing, a plethora of information is generated on genome and gene sequences, which is of paramount importance to understand gene function and the regulatory networks involved in plant growth, development, as well as stress tolerance at the molecular level. Plant genomics took a giant leap with the sequencing of the whole genome of Arabidopsis thaliana in 2000 (The Arabidopsis Genome Initiative 2000). This was followed by genome sequencing of important food crops like rice (Yu et al. 2002; Goff et al. 2002), sorghum (Paterson et al. 2009), maize (Schnable et al. 2009), barley (The International Barley Genome Sequencing Consortium 2012), pigeon pea (Varshney et al. 2012), chickpea (Varshney et al. 2013), and others. Rapid progress in plant genomics led to the discovery and isolation of important genes that regulate economically important traits and tolerance to biotic and abiotic stresses. With the availability of genome sequences in sorghum, the biggest challenge is to determine the functions of over 30,000 genes and deploy them in a practical genetic improvement program.

Being a C4 cereal and a drought-tolerant crop, sorghum always remained in the focus of genomics researchers. With the publication of sorghum genome sequence (Paterson et al. 2009), sorghum genomics has taken a paradigm shift and generated enormous information that has been integrated with other related crop species through comparative genomic studies. The completion of sequencing of sorghum genome and the creation of related genomics resources together with advances in the development of mapping populations and molecular marker resources have allowed researchers to accelerate the identification of agronomically important quantitative trait loci (QTLs) (Satish et al. 2009; Aruna et al. 2011; Mace et al. 2012; Nagaraja Reddy et al. 2013; Madhusudhana and Patil 2013). Advances in sequencing technologies such as next-generation sequencing (NGS) have resulted in sequence-based resources and related resource platforms for specific organisms including sorghum. Availability of whole-transcriptome profiling methods, advances in plant proteomics, simultaneous profiling of many metabolites, mutant populations, and biological databases have brought significant change in the approach in dealing with the biological processes. This chapter deals with the available genomics and bioinformatics resources in sorghum, which will help the sorghum researchers to develop useful information for the genetic improvement of sorghum.

2 Sequence Databases and Comparative Genomics Resources

Sorghum genome is comparatively small (~730 Mb), making it an attractive model for functional genomics of Saccharinae and other C4 grasses. It is the first C4 plant to be sequenced in 2009 by Andrew Paterson and his group involving 20 laboratories across the USA, Germany, China, Switzerland, and India. They observed that sorghum has ~75 % larger heterochromatin DNA as compared to rice. However, sorghum and rice have similar quantities of euchromatin. The net size expansion of the sorghum genome relative to rice predominantly involved long terminal repeat (LTR) retrotransposons. It was found that the sorghum genome contains 55 % retrotransposons, which is intermediate between rice (26 %) and maize (79 %) genomes. Paterson et al. (2009) modeled 34,496 sorghum genes, out of which ~27,640 were bona fide protein-coding genes. The accumulated information/data related to genome, transcriptome, proteome, and metabolome as a result of various high-throughput sequencing projects and proteomic and metabolomic studies are stored in different databases, which are summarized in Table 6.1. The comparative genomic databases like Gramene, PlantGDB, Phytozome, GreenPhylDB, CoGE, PLAZA, OrthologID, PlantTribes, SynBrowse, etc., along with associated web portals provide a uniform set of tools and automated analyses across a wider range of plant genomes. Among them, the first six resources deal with sorghum sequences along with other plant species. The salient features of these databases and their utility in comparative genomics are discussed below.

Table 6.1 Database resources for sorghum
Full size table

2.1 Gramene

Gramene (http://www.gramene.org/) is a curated, open-source, data resource for comparative genome analysis in the grasses developed in recognition of the importance of the grass family and put on public domain in 2002 from Cold Spring Harbor Laboratory (Ware et al. 2002). Initially, the rice sequence was used as base information to facilitate genomics research in other grass families like maize, sorghum, millet, sugarcane, wheat, oats, and barley. Subsequently, Ensembl Genomes at the European Bioinformatics Institute joined the group at the Cold Spring Harbor Laboratory, to further make this database as a resource for plant comparative genomics based on Ensembl technology. Besides rice, the database has information on barley, Brachypodium, foxtail millet, maize, oats, pearl millet, rye, wheat, and sorghum. In the recent version, information from other plant species like Glycine, Musa, Solanum, Brassica, Arabidopsis, Vitis, Populus, etc., have also been included. The goal of this database is to facilitate the study of cross-species homology relationships using information derived from public projects involved in genomic and EST sequencing, protein structure and function analysis, genetic and physical mapping, interpretation of biochemical pathways, gene and QTL localization, and descriptions of phenotypic characters and mutations. Even though a new version has been launched very recently, the information can be accessed through the old version also, which is organized in a better way. Information in the database organized in different modules (Fig. 6.1) are detailed as follows:

Fig. 6.1
figure 1

Organization of Gramene database (http://www.gramene.org/)

Full size image

“Genome” Module

This contains detailed information about the species, assembly, annotation, structural variation, besides references, and link to other species-related sites on the abovementioned plant species.

“Genetic Diversity” Module

This stores information on genotypes, phenotypes and their environments, germplasm, and association data. It also contains information from small-scale SSR diversity studies to large-scale SNP/InDel-based genotype-phenotype studies conducted in the mandated crops. With respect to sorghum, the database contains six datasets [Hamblin et al. (2004, 2006, 2007), White et al. (2004) and Casa et al. (2006)] and the results of the Sorghum Diversity Project.

“Pathways” Module

This contains four sub-modules, viz., RiceCyc, MaizeCyc, BrachyCyc, and SorghumCyc, which deals with information on pathway databases of the respective crops. It also provides mirrors of pathway databases from Arabidopsis, tomato, potato, pepper, coffee, Medicago, E. coli, and the MetaCyc and PlantCyc reference databases, thereby enabling comparative genome analysis. In this database, 297 pathways, 1,838 enzymatic reactions, and 9 transport reactions have been described. Known and/or predicted biochemical pathways and genes from sorghum are catalogued in SorghumCyc, which is primarily based on the genome annotations of Sorghum bicolor cv. BTx623. Many of the pathways might be incomplete or may contain errors since the functions of many of the sorghum genes are either provided by homology or HMM-based predictions.

“Protein” Module

This contains information on Swiss-Prot-TrEMBL protein entries from family Poaceae, which are annotated by the following three concepts of Gene Ontology (GO): (1) molecular function, (2) biological process in which it is involved, and (3) cellular component where it is localized. The associations assigned are based on annotations in the published literatures or generated through in silico approaches. Each association is supported with evidence (reference) and the evidence code (experiment type). On sorghum and related species, information on 35,817 proteins are available.

“Genes” or “Gene and Allele” Module

This contains detailed information on publicly available genes in cereal crops. Genes and their alleles associated with morphological, developmental, and agronomically important phenotypes, variants of physiological characters, biochemical functions, and isozymes are described here. Species-wise search for different gene types like “CDS, rRNA, tRNA, miRNA, siRNA, pseudogenes, not classified, sequenced gene loci or all gene types” is possible using wild cards.

“Ontologies” Module

This contains a collective information on controlled internationally accepted vocabularies and their associations to various objects such as QTL, phenotype, gene, proteins, and Ensembl rice genes for the following knowledge domains: Plant Ontology (PO), Trait Ontology (TO), Gene Ontology (GO), Environment Ontology (EO), and Gramene’s Taxonomy Ontology.

“Markers” Module

This contains the basic/primary information on the marker name, synonyms, source species, and a list of map positions of various markers used for mapping. This module has the link to “SSRIT tool,” which is useful for the identification of microsatellites.

“Maps” Module

This is primarily a visualization tool useful for visualizing the genetic, physical, sequence, and QTL maps for species dealt in the database. Comparative Map Viewer, referred as CMap, allows users to construct and compare different maps. All the data including the map sets, maps, features, and correspondences in this module are built from the “Markers” module. In CMap, the genomes of rice, sorghum, and Brachypodium are compared using syntenic blocks.

“QTL” Module

This contains quantitative trait loci (QTL) identified for numerous agronomic traits in the crops dealt in the database. Information on QTL along with associated traits and the mapped locus on the genetic map are available. With respect to sorghum, information on 136 QTL along with details of associated markers, linkage group, trait symbol, etc., are available.

“BLASTView” Module

This module provides an integrated platform for homology search against Ensembl plant databases, offering access to both BLAST (Basic Local Alignment Search Tool) and BLAT (BLAST-like Alignment Tool) programs. Species-wise search is possible in both DNA and protein databases using BLASTN (aligns the nucleotide sequences) and BLASTX (aligns translated sequences of any nucleotide sequence in all six reading frames), respectively.

“Gramene Mart” Module

This module has four databases, viz., Plant Gene 37, Plant variation 37, Gramene mapping, and Gramene QTL 37. Each database can be searched in 10 datasets of which sorghum is one.

“Species Page”

This contains detailed information on all the 11 cereal species dealt in the database with full phylogenetic information.

2.2 PlantGDB

PlantGDB was first reported by Dong et al. (2005), in which EST sequences were assembled into contigs that represent tentative unique genes. The functional annotation of these contigs was performed with the information derived from known protein sequences that were highly similar to the putative translation products. Initially, the database started with the data from only two plant species, viz., Arabidopsis and rice. Subsequently, PlantGDB (http://www.plantgdb.org) was published as a resource for comparative genomics across 14 plant species by Duvick et al. (2007). The aim of this web resource is to develop robust genome annotation methods, tools, and standard training sets for a number of sequenced or soon to be sequenced plant genomes. PlantGDB has four modules, viz., Sequence module, Genome module, Tools module, and Datasets module. Organization of the PlantGDB is shown in Fig. 6.2.

Fig. 6.2
figure 2

Organization of PlantGDB (http://www.plantgdb.org/)

Full size image

“Sequence” Module

This module can be used to BLAST search or to download nucleotide or protein sequences as well as access custom transcript assemblies. This module contains EST assemblies comprising PlantGDB-derived unique transcripts (PUT) assembled from plant mRNA sequences available at GenBank. Genome survey sequence (GSS) assemblies for maize and sorghum are also available.

“Genome” Module

This module contains genome sequence information on 16 dicots and seven monocots including sorghum (SbGDB). It has genome browsers to display current gene structure models and transcript evidence from spliced alignments of EST and cDNA sequences. The browsers also link community annotation tools to refine the gene annotations or to identify novel annotations. Each genome assembly is splice-aligned to transcripts as well as proteins from similar species and presented in a simple graphical interface (the xGDB platform).

“Tools” Module

This module provides a variety of tools for sequence analysis as follows:

  • BioExtract – a web interface to automate bioinformatics workflows. It is useful to query sequence databases, analyze data with bioinformatics tools, save results, and create and manage workflows.

  • Standard NCBI BLAST – useful to search against single or multiple BLAST databases simultaneously.

  • Distributed Annotation System (DAS) – useful to access PlantGDB annotations from the remote genome browsers.

  • GeneSeqer and GenomeThreader – useful to develop gene structure models based on spliced alignment to genomic sequences of both native and homologous ESTs, cDNAs, and protein sequences.

  • MuSeqBox – useful to examine multi-query sequence BLAST output, filter the BLAST hits based on user-defined criteria, and extract the informative parameters in tabular form.

  • PatternSearch – useful to search the specific patterns in genome sequence, i.e., short matches interspersed with mismatches and InDels.

  • ProbeMatch – allows the user to query his sequence against PLEXdb Probe Sequences.

  • TableMaker – an online search tool to access GenBank tables at PlantGDB using MySQL queries.

  • yrGATE – useful to create gene annotations in an xGDB genome browser itself. It shows all splice junctions revealed by EST/cDNA evidence and helps to create gene models and validate them.

“Datasets” Module

This module has datasets on AcDs Tagging Project, Alternative Splicing in Plants (ASIP) database, Plant Expression database (PLEXdb), Rescue-Mu tagged maize sequences, Maize-RFLP Full-Length Insert Sequencing Project, Splicing-Related Gene database (SRGD), and Uniform-Mu tagged maize sequences.

2.3 Phytozome

Phytozome (http://phytozome.jgi.doe.gov/pz/) is another online resource that was first released in 2008 to facilitate comparative genomic studies among green plants. It enables users with different computational abilities to access annotated plant gene families, navigate their evolutionary history, examine them in genomic context, assign putative function, and provide uniform access to complete genomes, gene and related sequences and alignments, gene functional information, and gene families, either as bulk information or as the result of user-defined queries (Goodstein et al. 2012). A number of commonly used open-source tools like Lucene, GBrowse (Stein et al. 2002), Jalview (Waterhouse et al. 2009), BioMart (Smedley et al. 2009), mView (Brown et al. 1998), and pygr are integrated in this portal which help in the gene family search, inspection, and evaluation. The Phytozome v7.0 contains data and analyses for 25 plant genomes, 18 of which are sequenced, assembled, and partially or completely annotated at the Joint Genome Institute (JGI). However, the recently released Phytozome v10 provides access to 47 sequenced and annotated green plant genomes including early-release genomes. With respect to sorghum, the initial release comprises the Sbi1 assembly and a Sbi1.4 gene set, which are the assembly and annotation reported by Paterson et al. (2009). Now, v2.1 is available as an early release as a result of modern annotation with additional RNA-seq data, comprising v2.0 assembly and v2.1 gene set. The genome is in 10 chromosomes with many short unmapped fragments; some may contain annotated genes (http://www.phytozome.net/sorghum_er.php).

Phytozome contains three important modules, viz., Species, Tools, and Info. “Species” and “Tools” modules have the options for keyword search, BLAST search, BLAT search, JBrowse, and bulk data. In addition to this, the “Tools” module has the options for the InterMine and BioMart, which is useful for data warehousing and construction of customized datasets with information on gene or gene families and annotation. With respect to poplar, Brachypodium, eucalyptus, and cassava genomes, both expression and diversity data can be viewed in JBrowse, searched, and downloaded from InterMine, as well as in bulk from the JGI Genome Portal. Screenshot of Phytozome database is depicted in Fig. 6.3.

Fig. 6.3
figure 3

Organization of Phytozome database (http://phytozome.jgi.doe.gov/pz/)

Full size image

Keyword and Sequence Similarity Search

Information on relevant attributes of gene and gene family like names, symbols, synonyms, external database identifiers, definitions, and functional annotation IDs can be retrieved by keyword search. BLAST and BLAT can be used to identify the genomic regions, gene transcripts, peptides, and gene families most similar to the query sequence. Gene families at a particular evolutionary node and families matching particular phylogenetic profiles can be searched. The database can also be searched for functional annotations to retrieve all matching functional identifiers and gene families.

Information on Gene and Gene Families

The Gene Family view gives information on each gene family and its constituent members. The default Genes in this family tab displays members of a particular gene family along with their source identifier, aliases, synonyms, and gene symbols. The family page has a set of lower tabs (Functional Annotation, MSA, and Family History) and upper tabs (Find related families, Align family members, Get Data, and Display options). The lower tab helps in the exploration of the evolutionary history of the gene family, while the upper tab is useful for analyzing the similarity among the related sub-families of genes. Besides depicting single gene functional annotations and evolutionary history, the Gene Page has links to alternatively spliced transcripts, if any; access to genomic, transcript, and peptide sequences associated with the gene; and a graphical view of other Phytozome peptides aligned against the peptide of the gene of interest.

Genome Browser

In GBrowse module, genome-centric views are provided for all the 41 genomes currently available in the Phytozome. This module can be accessed directly from the Phytozome home page, from individual member gene links available on the Gene Family or Gene Page, and from the BLAST/BLAT results page. Each browser shows a gene prediction track, where homologous peptides from related species, supporting ESTs, and one or more syntenic VISTA tracks identifying regions of this genome are depicted. The gene features and VISTA tracks are hyperlinked to the Gene Page and corresponding regions in the VISTA browser, respectively. On sorghum, 697,578,683 base pairs are arranged, which correspond to 34,496 loci and 36,338 protein-coding transcripts.

Data Retrieval

Bulk data files containing the genome assembly sequence, gene structure, transcript, coding, and peptide sequence in FASTA format and general annotation information are available for the genomes hosted in Phytozome database. Repeat-masked genome assemblies as well as supporting annotation data are also available for download. By using BioMart module, customized datasets can be constructed consisting of information on gene or gene family sequences and annotations based on user-defined data filters, attributes, and output formats. This module can be accessed from the “Get Data” tab on the Gene Family page or directly from the main menu of Phytozome.

2.4 GreenPhylDB

GreenPhylDB is a database specifically designed for comparative and functional genomics based on completely sequenced genomes. The development of GreenPhylDB v1.0 (5) by Conte et al. (2008) was inspired by the availability of whole-genome sequences of Arabidopsis thaliana and Oryza sativa genomes that offered opportunity for comparative genomics in plants. Since then, the most popular and reliable approach for the functional annotation of genes is by analyzing genes between species to identify orthologous genes (Kuzniar et al. 2008; Gabaldon et al. 2009). GreenPhylDB v2.0 was published during 2011 by Rouard et al. (2011) by adding 14 new genomes belonging to a major phylum of the plant kingdom including rodophytes, chlorophytes, mosses, lycophytes, and flowering plants with monocotyledons and dicotyledons. Six genomes were added in version 3, while 14 genomes were added in the current version (v4). Currently, the database has the genome information on 37 species. GreenPhylDB is accessible at http://www.greenphyl.org/cgi-bin/index.cgi.

The database represents a catalogue of gene families based on complete genomes. GreenPhylDB comprises complete proteome sequences from the major plant phylum, which are clustered to define a consistent and extensive set of homeomorphic plant families. Lists of plant- or species-specific gene families and several tools are provided to facilitate comparative genomics within plant genomes. The analyses include clustering of gene family followed by a phylogenomic analysis of the generated gene families. Upon validation of a cluster, phylogenetic analyses are performed to predict orthologs and ultraparalogs. Results of clustering are first manually annotated and then analyzed by a phylogenetic-based approach to predict orthologs, which is particularly useful for functional genomics and candidate gene identification of genes affecting agronomic traits of interests. This resource has 2,915 annotated gene families, of which 53 are specific to sorghum. Schematic diagram of the web resource is given in Fig. 6.4. GreenPhylDB has three important modules that can be used for comparative genomics analysis, viz., “Search” module, “Gene Family Lists” module, and “Tool Box” module.

Fig. 6.4
figure 4

Schematic diagram of GreenPhylDB (http://www.greenphyl.org/cgi-bin/index.cgi)

Full size image

“Search” Module

This module has the options of Quick Search and Advanced Family Search. The former searches based on text/keywords, while the latter on InterPro domains.

“Gene Family Lists” Module

This module contains the list of annotated gene families comprising 2,915 clusters and transcription factor families comprising 33 clusters. It has the option to list the gene families specific to a particular species/phylum. A GO Browser was developed as a web interface, which displays a list of terms defined in the Plant GO. By selecting a specific GO entry, the users can access a list of gene families potentially involved in plant growth and development along with the sub-classification of each identified gene family.

“Tool Box” Module

This contains the option for BLAST (sequence search with BLASTP or BLASTX), sequences to families (family classification of given sequence), Homolog sequences (get homologs and/or similar sequences with sequence ID inferred from phylogeny), InterPro domain (domain distribution is displayed by sequence and by species), Export sequences (provides a list of sequence ID used in database that can be exported in the selected format), TreePattern (to explore phylogenetic trees), and Create family (to create gene families). It has 32,796 sorghum-specific proteome datasets in it. The biggest advantage of this portal is to construct gene family tree and identify homologs.

2.5 CoGe

CoGE stands for Accelerating Comparative Genomics (https://genomevolution.org/CoGe/). It is a unique web resource having many interconnected tools to create open-ended analysis networks. The features of CoGe were published by Lyons and Freeling (2008) and Lyons et al. (2008). CoGe is designed to address four issues: (1) single platform to store multiple versions of multiple genomes from multiple organisms, (2) rapid identification of sequences of interest in genomes of interest (with associated information), (3) comparison of multiple genomic regions using any algorithms, and (4) visualization of the results for easy and quick identification of “interesting” patterns. Organization of CoGe database is given in Fig. 6.5. CoGe database has a set of tools for comparative genome analysis. They are as follows:

Fig. 6.5
figure 5

Organization of CoGe database (https://genomevolution.org/CoGe/)

Full size image

  • OrganismView – searches and gives an overview of an organism and its genomic information

  • CoGeBlast – BLAST sequences against any number of organisms of user’s choice

  • FeatView – searches for genomic features by name or description

  • SynMap – generates syntenic dotplots of any two genomes

  • SynFind – identifies syntenic regions across many genomes

  • GEvo – compares multiple genomic regions using a variety of sequence comparison algorithms for high-resolution analysis to quickly identify patterns of genome evolution

An Integrative Orthology Viewer combines information from different orthology prediction methodologies. Central tools and access points of CoGe allow to find sequences of interest, and “hub” points direct from one part of the system to another. For example, if a region with an inversion is identified during the comparison of sorghum with the maize genome using SynMap, breakpoints of that region may be compared using GEvo in high detail, and the maize sequence can be extracted out using SeqView. Subsequently, FeatView can be used to identify all the protein-coding regions, and the information generated can be used to find homologs in other plant genomes using CoGeBlast. GEvo can be used to validate putative syntenic regions. If, say, a gene with extra copy number is identified in a syntenic region, its sequence may be obtained using FeatView once again. Putative intra- and interspecific homologs of it may be obtained using CoGeBlast, which will generate a FASTA file using FastaView. This can be aligned using CoGeAlign and used to build a phylogenetic tree through TreeView or exported to more expansive phylogenetic platform such as CIPRES. Simultaneously, the codon and protein usage variation of the genes may be checked using FeatList. If some interesting variation is observed in some genes, their overall GC content and wobble-position GC content may be checked in FeatView. Horizontal transfer of DNA fragments/genes from the mitochondria can be identified using CoGeBlast or GEvo.

2.6 PLAZA

A centralized plant genomics platform is essential for performing evolutionary and comparative analyses of gene families and genome organization, which integrates all the information generated by various sequencing projects along with advanced tools for data mining. PLAZA is a versatile plant comparative genomics resource centralizing genomic data from different genome sequencing initiatives (http://bioinformatics.psb.ugent.be/plaza/) published by Proost et al. (2009). Plant sequence data and comparative genomics methodologies are integrated in an online platform with interactive tools to study gene function and gene and genome evolution within the green plant lineage. It has integrated structural and functional annotation of 25 green plant species, which includes 909,850 genes. Out of these genes, 85.8 % are protein coding, which are clustered in 32,294 multigene families, resulting in 18,547 phylogenic trees. In addition to the basic information related to gene structure and function such as genome coordinates, mRNA and protein sequences, and gene description, PLAZA offers various tools to browse genomic data for homology, ranging from local synteny to gene-based colinearity views useful for comparative genomics plant genome evolution. Organization of PLAZA database is given in Fig. 6.6.

Fig. 6.6
figure 6

Organization of PLAZA database (http://bioinformatics.psb.ugent.be/plaza/)

Full size image

  • Synteny plot – a basic tool that shows all genes from the specified gene family along with their neighboring genes, thereby helping to study genomic homology in comparison to colinearity.

  • WGDotplot – useful for analyzing genome-wide colinearity leading to the identification of large-scale duplications or to study genomic rearrangements within or between species.

  • Skyline plot – useful for browsing multiple homologous genomic segments and provides a comprehensive view of the regions that are colinear in the species selected by the user.

  • Workbench – useful to analyze multiple genes in batch that are uploaded through gene identifiers or based on similarity search and calculate different genome statistics for user-defined gene sets.

  • Whole Genome Mapping tool – useful to display a selection of genes on the chromosomes and to view the distribution of different classes of genes such as protein coding, pseudogene, or transposable element. It also provides information about the gene duplication.

  • Advanced query system – useful for the rapid retrieval of relevant information using different data types and research tools.

2.7 CSGRqtl

CSGRqtl is a comparative genomic database (http://helos.pgml.uga.edu/qtl/) developed by Zhang et al. (2013a) as a data mining resource specific to crops, weeds, and models of Saccharinae clade. This database complements and supplements the database Gramene, which contains mapping data from a wide spectrum of grasses. CSGRqtl uses sorghum genome sequence as a reference with an aim of anchoring published QTLs of the Saccharinae clade to the sorghum genome. This database uses the Plant Trait Ontology defined by Gramene and facilitates the data comparisons among the grasses of Saccharinae clade and between Saccharinae and other taxa to categorize quantitative trait loci (QTLs) with their approximate physical positions. CSGRqtl also integrates gene annotations, genetic markers, and paleoduplicated regions, which facilitate QTL mapping and the study of candidate gene underlying the QTLs. Organization of CSGRqtl database is given in Fig. 6.7.

Fig. 6.7
figure 7

Organization of CSGRqtl database (http://helos.pgml.uga.edu/qtl/)

Full size image

CSGRqtl is equipped with a number of tools for analysis, such as text-based search, trait ontology browser, QTL correspondence, and CMap database, to allow a user to query and visualize the background database.

Text-Based Search

QTL search based on the trait of interest gives a set of QTLs underlying the trait. Genome-wide overview of QTL distribution is depicted by a circular plot created by the bioinformatics resource Circos. It also identifies the potential QTL hot spots in the genome. QTL search based on a sorghum gene identifier or annotation gives a list of QTLs containing the queried gene, and the approximate positions of genes and QTLs are depicted by a plot.

Trait Ontology Browser

The trait ontology browser displays the hierarchy of trait ontology and lists QTLs associated with each trait accession since each QTL is allotted a trait accession defined by Gramene Plant Trait Ontology.

QTL Correspondence

The associations between paleoduplicated regions and QTLs in rice and sorghum are depicted by circular plots to indicate non-overlapping QTLs divided by inter-genomic synteny or narrowed by intra-genomic synteny. The users can download orthologs/paralogs for genes associated with non-overlapping QTLs for the trait of interest.

CMap Database

Alignments between genetic maps and the sorghum genome sequence can be viewed by the user. It also gives information on RFLP probe and SSR primer sequences for each anchored marker that is amenable for alignment.

Genome Browser

This is implemented using Generic Genome Browser version 2.39 (Stein et al. 2002) and used to associate sorghum QTL data with gene annotations. This browser contains gene models from standard sorghum genome annotation version 1.4 and about 209,828 sorghum ESTs from the NCBI. It also contains GC content, six-frame translation, and restriction sites for each genomic region. The user can get information of all annotated genes for a particular QTL region and also can access all QTLs in any genomic region.

2.8 SorGSD

DNA sequence variations between diverse sorghum lines are an important pre-requisite for the genetic improvement of sorghum for agronomic traits as well as tolerance to biotic and abiotic stresses through breeding by design and high-efficiency genomic selection. Advances in the next-generation sequencing (NGS) technologies have brought about a surge in the re-sequencing of the diverse sorghum accessions belonging to different categories such as improved inbreds, landraces, wild/weedy sorghums, and wild relatives. Recently, a diverse panel of 48 sorghum accessions which were divided into four groups, including improved inbreds, landraces, wild/weedy sorghums, and a wild relative Sorghum propinquum, has been re-sequenced leading to the generation of enormous amount of SNP data (Mace et al. 2013). Proper organization of this SNP data will offer excellent opportunity for researchers to identify variation in their genes of interest, explore evolutionary relationships among cultivated and wild types, develop DNA markers for future genetic studies, and utilize this data for genome-wide association studies (GWAS). With this premise, SorGSD, a web-based large-scale genome variation database, was developed during August 2014 and maintained by the Data Management Center, Beijing Institute of Genomics (BIG), Chinese Academy of Sciences, and the Laboratory for Conservation and Utilization of Bio-resources, Institute of Botany, Chinese Academy of Sciences. The database contains 62 million SNPs with annotations assisted by an easy-to-use web interface for users for efficient browsing, searching, and analysis of the SNPs. The pipeline for SNP calling included trimming of adapter and filtering of all low-quality reads, the use of BWA (version 0.6.2-r126) to map clean read to sorghum reference sequence (V1.4), SAMtools package to convert mapping results to BAM format, Picard (version 1.87) program to eliminate duplicated reads generated during the process of library construction, SNP calling by GATK (version 2.5-2-gf57256b) toolkit, SNP identification based on the quality estimation scores generated by GATK (quality value ≥30 and depth of coverage ≥5), and SnpEff program for the annotation of SNPs. This database is a rich repository to molecular breeders for the identification of biomarker, genetic analysis, and marker-assisted breeding of sorghum and other crops. The SorGSD can be accessed from http://sorgsd.big.ac.cn/snp/. The SorGSD has four modules, viz., Browse, Search, Compare, and Download as shown in Fig. 6.8.

Fig. 6.8
figure 8

Organization of SorGSD database (http://sorgsd.big.ac.cn/snp/)

Full size image

Browse Module

This module can be used to browse total SNPs as well as gene-wise and chromosome-wise SNPs. SNPs in Gene lists SNPs located in gene and coding regions. The “Coding,” “Synonymous,” and “Non-synonymous” lists are used to view SNP located in coding region, annotated as synonymous and non-synonymous. Query can be given based on the chromosome number also. The users can browse SNP information and their relevant annotations for each sorghum line. The database uses GBrowse to visualize InDel, SNP, gene, transcript, density information of SNP/300 kb, and allele frequency.

Search Module

This module helps in searching SNPs in a single individual by setting parameters such as chromosome location, SNP class, and SNP location in gene region and genotype. Options are provided for the selection of SNP annotation and SNP genotype. The search results can be either visualized graphically in a genome browser or displayed in formatted tables.

Compare Module

This module helps in comparing SNPs in two or more individuals by setting parameters such as chromosome location, SNP class, and SNP location in gene region and genotype. The SNPs in selected individuals can be compared with that of a single reference genotype or more genotypes. Options are provided for the selection of SNP annotation and SNP genotype.

Download Module

This module contains the datasets, viz., SNP files, InDel files, SRA files, and Fastq files, which can be directly downloaded for further analysis.

3 Genomics Resources for Analyzing DNA Sequence Variation

In the current era of genomics, large-scale EST as well as genome sequencing projects resulted in the generation of enormous amount of DNA sequence data that are organized and stored in various databases. Such data available in the public domain are the main targets for the development of molecular markers such as simple sequence repeats (SSRs), insertion-deletions (InDels), single nucleotide polymorphisms (SNPs), etc., through various computational approaches and bioinformatics tools available. These molecular markers are the best tools available for the plant geneticists for analyzing the DNA sequence variations through an easy and rapid PCR assay and are useful for assessing the genetic diversity and population structure of germplasm lines, varietal identification, genetic purity testing of hybrids and parental lines, mapping of genetic loci through QTL mapping, and marker-assisted selection.

Assessment of genetic diversity in the germplasm accessions or the parental line gene pool is the primary step in any plant breeding program. Earlier, morphological as well as quantitative traits were used for the assessment of genetic diversity. Due to their inherent limitations in the number as well as environmental influence, molecular markers have become the choice of such genetic diversity assessments since these markers are environmentally neutral. Several studies were undertaken over the years in the assessment of genetic diversity using molecular markers such as RAPD (Ayana et al. 2000; Uptmoor et al. 2003), RFLP (Tao et al. 1993; Ahnert et al. 1996), AFLP (Geleta et al. 2006; Ritter et al. 2007), and ISSR (Aruna et al. 2012). These markers are also used in the mapping of major genes (Knoll et al. 2008; McIntyre et al. 2008) as well as QTL (Srinivas et al. 2009b; Satish et al. 2009).

SSR markers are the widely used PCR-based markers for various genetics and mapping studies in sorghum due to their abundance in the genome, highly polymorphic nature, and easy assay. Prior to the completion of genome sequencing of sorghum in 2009, several research groups have developed a large number of SSR markers which were subsequently used for various genetic studies in sorghum. These studies helped in the development of genomic SSR markers [Xtxp series (Kong et al. 2000; Bhattramakki et al. 2000, http://sorgblast3.tamu.edu/search/marker.htm), XSb series (Taramino et al. 1997), Xgap series (Brown et al. 1996)], SSR markers derived from cDNAs [Xcup series (Schloss et al. 2002)], whole-genome sequence-based SSR markers [SB series (Yonemaru et al. 2009)], expressed sequence tag (EST)-based SSR markers [Xisep series (Ramu et al. 2009), Xiabt series (Arun 2006; Reddy et al. 2008), Stgnhsbm and Dsenhsbm series (Srinivas et al. 2008, 2009a)], SSR markers derived from unigenes [Ungnhsbm series (Srinivas et al. 2009b; Nagaraja Reddy et al. 2012)], (GATA)n motif-based SSR markers [SbGM series (Jaikishan et al. 2013)], and other SSR markers of an unknown type [gpsb and mSbCIR series (developed at CIRAD, France, and partially published in Mace et al. 2009)]. The details of the SSR markers developed by different sorghum groups are given in Table 6.2.

Table 6.2 SSR markers developed by different sorghum research groups
Full size table

Even though several studies on the assessment of genetic diversity were reported over the years, very few studies have done a comprehensive analysis and resulted in a set of robust SSR markers that can be used universally across laboratories for this purpose. A set of 38 SSR markers distributed across 10 chromosomes of sorghum selected based on three different linkage maps were used to establish the diversity research set comprising 107 sorghum accessions (Shehzad et al. 2009) from a set of 320 sorghum germplasm accessions. A diversity analysis kit was developed (Billot et al. 2012), which contains information on 48 robust sorghum SSR markers that can be used to calibrate SSR genotyping data acquired with different technologies and compare them to genetic diversity references. A reference set comprising a wide range of sorghum genetic diversity was screened with 40 EST-SSR markers by Ramu et al. (2013), and the analysis highlighted the greater discriminating power of these markers as compared to the genomic SSR markers.

In the current era of genome sequencing, the discovery of SNPs and insertion and (or) deletions (InDels) through high-throughput methods has led to a revolution in their use as DNA markers (Batley and Edwards 2007; Batley et al. 2007; Edwards et al. 2007). Advancements in sequencing technologies, execution of re-sequencing projects, and availability of the enormous amount of ESTs along with the development of efficient computational platforms have helped in the rapid discovery of SNPs and InDels in sorghum. SNPs may be considered the ultimate genetic marker as they represent the finest resolution of a DNA sequence, generally are abundant in populations, and have a low mutation rate (Syvanen 2001). The mining of readily available sequence data for SNPs through in silico approaches significantly reduces the costs (Taillon-Miller et al. 1998), and several SNP mining tools have been developed (Barker et al. 2003; Batley et al. 2003; Savage et al. 2005; Chagne et al. 2007). Sorghum researchers across the globe have utilized different types of data, such as ESTs, whole-genome re-sequencing data, and genotyping-by-sequencing data for the discovery of SNPs with the help of various computational tools. The detail of the SNPs developed by different sorghum groups is given in Table 6.3.

Table 6.3 SNPs developed by different sorghum research groups
Full size table

InDels are next only to SNPs in terms of their abundance in the genome. However, InDels can be converted into PCR-based markers and can be resolved through routine gel electrophoresis systems. InDels exhibit length polymorphisms that have been successfully exploited in sorghum in the mapping of important loci such as waxy (McIntyre et al. 2008) and tannin (Wu et al. 2012). In sorghum, about 99,948 InDels of 1 to 10 bp in length were detected by Zheng et al. (2011) through a genome-wide analysis in sweet and grain sorghum. Potential intron polymorphism (PIP) markers developed by Yang et al. (2007) are a unique type of markers that targets both the SNPs and InDels. Among the two types of polymorphisms, intron length polymorphism (ILP) can be easily detected by exon-primed intron-crossing PCR (EPIC-PCR) (Palumbi 1995), where primers are designed in exonic regions flanking the target introns. Potential intron polymorphism (PIP) database for plants was developed by Yang et al. (2007) comprising a total of 57,658 PIP markers for 59 plant species, of which 4314 are of sorghum. These markers can be exploited for genetic diversity assessment, cultivar identification, mapping, and marker-assisted selection.

A new high-throughput hybridization-based marker technology that could serve as an efficient alternative to low-throughput gel-based marker systems was reported in sorghum by Mace et al. (2008). This system does not require sequence information and is amenable for high multiplexing. A genotyping array was developed with ~12,000 genomic clones using PstI + BanII complexity with a subset of clones obtained through the suppression subtractive hybridization (SSH) method. About 508 markers were polymorphic and were used for the genetic diversity analysis of 90 diverse sorghum genotypes and for the construction of a genetic linkage map for a cross between R931945-2-2 and IS 8525. These markers are useful for whole-genome profiling and can be used for diversity analyses and construction of medium-density genetic linkage maps.

A robust SNP array platform was developed recently by Bekele et al. (2013) using 2,124 selected Infinium Type II SNPs from a total of over one million high-quality SNPs identified by the alignment of whole-genome sequences (6–12× coverage) of genetically diverse genotypes comprising two grain and three sweet sorghum genotypes of S. bicolor and an additional 876 SNPs selected based on their phenotypic association with early-stage chilling tolerance identified by phenotype-based pool sequencing. Testing this array with selected SNPs using 564 genotypes comprising four unrelated RIL and F2 populations and a genetic diversity collection resulted in the validation of 2,620 robust and polymorphic SNPs. This SNP array platform is very useful for genetic mapping, genome-wide association, and genomic selection.

4 Bioinformatics Resources for DNA Sequence Analysis

DNA sequence variations arise either due to point or gross mutations. Point mutations are mostly due to base substitution (transition or transversion). Gross mutation may be insertion or deletion (InDels) of few to large sequences. It also may arise by duplication, inversion, and translocation. Gross mutations involving large sequences can easily be detected cytologically, while it is difficult to detect gross mutations of smaller dimension (say < 500 bases). Sequencing helps us to detect this variation very precisely. For this purpose, the mutant sequence is aligned with the wild-type sequence, and sequence variation is detected.

Sequence alignment is a way of arranging the nucleic acid (DNA, RNA) or protein sequences to identify regions of similarity that may be a result of functional, structural, or evolutionary relationships between the sequences (Mount 2004). Aligned nucleotide or amino acid sequences are generally represented as a row matrix by inserting gaps between the residues such that identical characters are aligned in successive columns. In the case of two sequences sharing a common ancestor, mismatches are interpreted as point mutations and gaps as InDels. Conservation of base pairs indicates a similar functional or structural role of the target sequence. Very short or very similar sequences can be aligned manually. However, with sequencing projects, large sequences are available, and these need to be aligned in large number, which is not possible manually. Different algorithms have been developed to facilitate high-quality sequence alignments. These include dynamic programming, heuristic algorithms, or probabilistic methods. Computational approaches to sequence alignment are of two categories: global alignments and local alignments.

Global Alignment

With this approach, both sequences are aligned along their entire lengths, including every nucleotide or amino acid, and the best alignment is found. This approach is most useful if the query sequences are similar and of roughly equal size. Dynamic programming called Needleman-Wunsch algorithm is very popularly used for this purpose.

Local Alignment

In this approach, the best subsequence alignment, which includes only the most similar sequence, is found. This approach is useful, particularly for dissimilar sequences that are expected to contain similar sequence motifs within their larger sequence. The dynamic programming method, namely, Smith-Waterman algorithm, is commonly used for this purpose.

Both the global and local alignments lead to erroneous conclusion when the downstream part of one sequence overlaps with the upstream part of the other sequence. In such situations, hybrid methods such as “semi-global” or “glocal” (short for global-local) methods are employed. These methods help in finding the best possible alignment that includes the start and end of one or the other sequence. The sequence alignment may be of two types based on the number of sequences used for alignment, viz., pair-wise sequence alignment (PSA) and multiple sequence alignment (MSA).

PSA

This is the comparison of two biological sequences (nucleic acid or protein) at a time to reveal the similarity or homology between them. This helps in finding the best-matching local or global alignments of two query sequences. This method is most useful in situations where extreme precision like searching of database for sequences with high similarity is not required. Dot matrix methods, dynamic programming, and word methods are mostly used for pair-wise alignments. Several PSA tools have been developed by several workers (Table 6.4), many of them are free to use, and some of them are available as commercial software. BLAST is the best example for pair-wise sequence alignment. BLAST searches a query sequence (discovered sequence) against a database of known sequences in order to find similarities. The program compares nucleotide or protein sequences to sequence databases and calculates the statistical significance of matches. BLAST searching can be performed using the web-based applications (Web BLAST) or can be run locally in the PC provided it has an existing database to aid searching. There are five types of BLAST:

Table 6.4 Important resources for pair-wise alignment
Full size table

  • BLASTN: Compares a nucleotide query sequence against a nucleotide sequence database

  • BLASTP: Compares an amino acid query sequence with a protein database

  • BLASTX: Translates a DNA sequence into six protein sequences using all six possible reading frames and then compares each of these proteins to protein database

  • TBLASTN: Translates every DNA sequence in a database into six potential proteins and then compares the protein query against each of those translated proteins

  • TBLASTX: Translates DNA from both a query and a database into six potential proteins and then performs 36 protein-protein database searches

MSA

This is the comparison of more number of sequences (three or more) simultaneously. It helps in predicting the structure and function of a given protein sequence due to its ability to detect common features and conserved domains across the sequences analyzed. It also helps to discover novel and related sequences and to construct and search for sequence patterns. Detection of conserved regions among homologous sequences is useful in designing PCR primers. There are a number of MSA programs available in the public domain (Table 6.5).

Table 6.5 Important resources for multiple sequence alignment
Full size table

Clustal is a widely used multiple sequence alignment tool (Chenna et al. 2003). There are three main variations: ClustalW (command line interface, Larkin et al. 2007), ClustalX (this version has a graphical user interface, Thompson et al. 1997), and Clustal Omega [(allows hundreds of thousands of sequences to be aligned in only a few hours). It will also make use of multiple processors, where present. In addition, the quality of alignments is superior to previous versions (Sievers et al. 2011)]. A wide range of input formats, including NBRF/PIR, FASTA, EMBL/Swiss-Prot, Clustal, GCC/MSF, GCG9 RSF, and GDE, are acceptable in this program. The output format can be one or many of the following: Clustal, NBRF/PIR, GCG/MSF, PHYLIP, GDE, or NEXUS. There are three main steps in the alignment process, i.e., pair-wise alignment, followed by the creation of a guide tree (or use a user-defined tree) and finally the use of the guide tree to carry out a multiple alignment. If “Do Complete Alignment” option is selected, all these steps are done automatically, or else the task may be carried out following options, viz., “Do Alignment from guide tree” and “Produce guide tree only.” There is an option for default setting or customized settings.

Alignments may be represented graphically and in text format. An asterisk or pipe symbol is commonly used to show identity between two columns. Colors are also used by several programs to display identity and dissimilarity. The sequence alignment results are stored in a variety of text-based file formats. Most common input and output formats are FASTA format and GenBank format.

5 Resources for Analyzing Transcriptomes

Transcriptome analysis involves the screening of candidate genes, predicting its function, and discovery of regulatory elements through high-throughput gene expression analysis. Initially, large-scale sequencing of ESTs was used as the main approach for transcriptome analysis. Later, the hybridization-based methods such as microarrays/GeneChips were developed and popularly used for large-scale gene expression analysis. Sequencing-based methods such as serial analysis of gene expression (SAGE) and massively parallel signature sequencing (MPSS) have been successfully employed for the identification of a large number of transcripts along with quantitative comparison of transcriptomes (Velculescu et al. 1995; Brenner et al. 2000). Furthermore, as a next-generation DNA sequencing application, deep sequencing of short fragments of expressed RNAs, including sRNAs, is quickly becoming an efficient tool for use with genome-sequenced species (Harbers and Carninci 2005; de Hoon and Hayashizaki 2008).

A sorghum cDNA microarray providing data on 12,982 unique gene clusters was used by Buchanan et al. (2005) to examine genome-wide changes in gene expression in sorghum seedlings under high salinity (150 mM NaCl), osmotic stress (20 % polyethylene glycol), or abscisic acid (125 μM ABA). A total of 3,508 cDNAs selected from the two cDNA libraries constructed from a strong greenbug resistance sorghum line (M627) and a susceptible line (Tx7000) with or without infestation were used to develop a cDNA microarray for the identification of sorghum genes responsive to greenbugs (Park et al. 2006).

An Agilent rice gene expression microarray (product number: G2519F, 44 K) was used to study tissue-specific gene expression profiles of S. propinquum with special emphasis on rhizome development by Zhang et al. (Zhang et al. 2013b) that contained 45,220 independent probes (60-mer) corresponding to 21,495 O. sativa mRNA sequences available in GenBank due to the non-availability of microarray platform in sorghum. Only recently, the first whole-transcriptome microarray was developed in sorghum by Shakoor et al. (2013) comprising a gene chip containing 1,026,373 probes covering 149,182 exons (27,577 genes) across the nuclear, chloroplast, and mitochondrial genome along with putative non-coding RNAs to identify tissue-specific genes and novel regulatory sequences. Toward identification and functional characterization of genes in sorghum genome, Shakoor et al. (2014) used the first commercial whole-transcriptome sorghum microarray chip (Sorgh-WTa520972F) to identify tissue- and genotype-specific expression patterns using grain, sweet, and bioenergy sorghums. Microarray dataset was generated using 78 samples involving different tissue types (shoot, root, leaf, and stem) and dissected stem tissues (pith and rind) of six diverse genotypes (R159, Atlas, Fremont, PI152611, AR2400, and PI455230), which revealed tissue- and genotype-specific expression patterns of different metabolic pathways indicating the importance of intraspecies variations in sorghum.

Next-generation sequencing technologies have enabled the researchers to characterize small RNA component of the transcriptomes in many plant species. According to the recent release of the miRBase database (http://www.mirbase.org, release 20: June 2013), about 205 miRNAs are described for sorghum, whereas 592 miRNAs are described for rice. The sorghum genome sequencing consortium identified 149 predicted miRNAs belonging to 27 miRNA families (Paterson et al. 2009). The identification of miRNAs from different target tissues, developmental stages, and stress treatments offer an excellent opportunity to understand the role of miRNAs in the regulation of expression of genes influencing traits of agronomic importance.

Prior to whole-transcriptome microRNA (miRNA) sequencing projects, computational approaches based on homology search were used for the identification of miRNAs in different plant species. Using this approach, Du et al. (2010) identified a total of 17 new miRNAs based on the GSS and the miRNA secondary structure that were distributed unevenly among 11 miRNA families. Analysis of these miRNAs via online software miRU revealed that they might regulate 64 target genes, most of which are involved in RNA processing, metabolism, cell cycle, protein degradation, stress response, and transportation. The small RNA component of the transcriptome of grain and sweet sorghum stems was characterized by Calviño et al. (2011) using F2 population derived from the cross between BTx623 (grain sorghum) and Rio (sweet sorghum) that segregated for sugar content and flowering time. They reported that the variation in miR172 and miR395 expression correlated with flowering time, while that of miR169 correlated with sugar content in stems.

With the increasing number of whole genomes, large-scale cDNA sequencing, and microarray projects in the recent years, an enormous amount of transcriptome data is generated and stored in public databases. This data serves as a valuable resource for many secondary uses, such as co-expression and comparative transcriptome analyses. NCBI’s Gene Expression Omnibus (GEO) (http://www.ncbi.nlm.nih.gov/geo/) and the European Bioinformatics Institute (EBI)’s ArrayExpress (http://www.ebi.ac.uk/arrayexpress/) are similar such databases, which serve as the primary archives of transcriptome data in the public domain (Parkinson et al. 2007; Barrett et al. 2009). Transcriptome data of S. propinquum in relation to rhizome development and the data of grain (BTx623) and sweet (Keller) sorghum are stored in GEO. ArrayExpress database has the transcriptome datasets of tissue-specific transcriptomic profiling of S. propinquum using a rice genome array, sorghum gene expression using Agilent custom 4x44K microarray, RNA-Seq of S. bicolor 9d seedlings in response to osmotic stress and abscisic acid, and high-throughput sequencing of small RNAs in S. bicolor. Another database, namely, EGENES (http://www.genome.jp/kegg-bin/create_kegg_menu?category=plants_egenes), is a multi-species resource integrating the genomic, chemical, and network information comprising genes, molecules, and biological pathways representing the cellular functions. This resource is useful for the comparison and mutual validation of genome-based pathway annotation and EST-based annotation. The EGENES consists of data on 25 eukaryotic species including sorghum. The sorghum datasets include 190,946 ESTs, 19,597 contigs, 23,171 singletons, 122 pathway maps, and 1,189 mapped contigs.

6 Genetic Resources for Mapping Agronomically Important Traits

The majority of the agriculturally important traits such as yield, biotic and abiotic stress tolerance is complex and is governed by quantitative trait loci (QTLs). These QTLs are influenced by the environment and the interaction between QTL and environment. Linkage mapping (biparental mapping) and linkage disequilibrium (LD) mapping (association mapping) are the two most commonly used approaches for the dissection of such complex traits. The first step in QTL mapping is the development of mapping populations by crossing parental lines that are contrasting for the trait of interest (biparental population and multi-parental population) or assembling of diverse germplasm lines (natural population), which serve as an important genetic resource.

Biparental Population

This approach involves the mating between two parental lines that are contrasting for the trait of interest and advancing them to develop F2, backcross populations, recombinant inbred lines (RILs), backcross inbred lines (BILs), and double haploid (DH). In addition to this, introgression lines (ILs) and near isogenic lines (NILs) are also developed. These mapping populations are known as a first-generation mapping resource. Even though several preliminary studies used F2 populations, the use of advanced generations, particularly RILs derived by single-seed descent from F2 individuals from a cross between two distinct homozygotes, is most commonly used for QTL mapping purposes (Keurentjes et al. 2011) because they are immortal and can be multiplied any number of times (Huang et al. 2011) for phenotyping in different environments/seasons. To tackle the problem of epistasis due to the interaction between multiple loci, the biparental populations such as ILs and NILs are also used (Rakshit et al. 2012). The advantage of employing NIL over RIL is mainly the detection of minor QTL that are missed while using RILs (Keurentjes et al. 2007). Globally, many sorghum research groups have developed and used several RIL populations for various traits (Table 6.6).

Table 6.6 Biparental mapping populations in sorghum
Full size table

Multi-parental Mapping Populations

Biparental populations, though popularly used for QTL mapping, have two major limitations such as relying on the recombination events happening in the F1 and subsequent generation and mapping only the allelic pairs that are present in the two contrasting parents (Rakshit et al. 2012). This affects the map resolution of the QTL since QTL will be placed on a large chromosomal region (Li et al. 2010). In the recent past, the second-generation mapping resources such as association mapping, nested association mapping (NAM), and multi-parent advanced generation inter-cross (MAGIC) were developed to overcome the limitations of biparental populations. NAM populations are developed by crossing a central parent with other diverse parents in a star design (Huang et al. 2011), and such populations have been established in maize (Yu et al. 2008; Buckler et al. 2009; McMullen et al. 2009) and Arabidopsis (Bentsink et al. 2010; Brachi et al. 2010). Development of a large NAM population in sorghum was reported recently by Jordan et al. (2012) comprising more than 4,000 lines from 100 sub-populations derived from a large BC1F1 population using a single elite line as the recurrent parent resulting in the sampling of the diversity of sorghum including wild relatives. Such populations help in fine mapping of QTL; however, the interaction of QTL with genetic background cannot be analyzed since one parent is common in all sub-populations. Consequently, the concept of MAGIC population was proposed by Cavanagh et al. (2008) to address the major limitations of biparental mapping populations. This concept was used as an additional resource for dissecting the genetics of natural varieties in Arabidopsis multi-parent recombinant inbred line (AMPRIL) population (Huang et al. 2011). Recently, another multi-parent mapping population known as wide diallel population derived from 19 founder lines of sorghum selected from a wide gene pool was used to map the heterotic trait locus and to identify intra-locus interactions underlying hybrid vigor (Ben-Israel et al. 2012).

Natural Populations

Linkage analysis involves analyzing a limited number of recombination events that occur during the construction of mapping populations, which results in the localization of QTL in the interval of 10–20 cM. Moreover, cost is involved in the propagation and evaluation of a large number of lines (Doerge 2002; Holland 2007). While several linkage analysis studies have been conducted in sorghum over the past two decades, only a limited number of genes were cloned or tagged at the gene level (Bout and Vermerris 2003; Saballos et al. 2009, 2012). Association mapping, also known as LD mapping, has emerged as an important tool to dissect the genetics of complex traits at the sequence level by exploiting the recombination events accumulated in the natural population (germplasm lines) during the course of evolution (Nordborg and Tavaré 2002; Risch and Merikangas 1996). According to Yu and Buckler (2006), association mapping has three advantages as compared to conventional linkage analysis, viz., (1) better mapping resolution, (2) reduction in research time, and (3) access to greater allele number. In addition, association mapping enables researchers to use next-generation sequencing technologies to exploit the diversity present in the natural populations, the value of which is known to crop breeders but exploited to a limited extent.

Sorghum is most suitable for association mapping of complex traits since it harbors one fourth sequence diversity that of maize (Hamblin et al. 2006), a 26-fold less population recombination than in maize (Hamblin et al. 2005), and natural homozygosity. LD is extensive enough in sorghum, which allows the simplification of a large number of SNPs into a smaller number of haplotypes resulting in reduced genotyping costs and increased statistical power (Clark 2004). Association mapping studies in sorghum using natural populations have been reported for plant growth and development (Kong et al. 2013), plant height (Murray et al. 2009; Wang et al. 2012), grain quality (de Alencar Figueiredo et al. 2010; Sukumaran et al. 2012), morphological traits (Shehzad et al. 2009), Brix (Murray et al. 2009), kernel weight and tiller number (Upadhyaya et al. 2012a), plant height and maturity (Upadhyaya et al. 2012b), endosperm quality (Hamblin et al. 2007), and agroclimatic traits (Morris et al. 2013).

Assembly of the association mapping panel comprising sorghum germplasm accessions possessing extensive genetic diversity is an important prerequisite for any association mapping study. Few association mapping panels were developed by different sorghum research groups across the world. An association mapping panel comprising 377 accessions representing all major cultivated races and important US breeding lines along with their progenitors was assembled and characterized for genetic and phenotypic diversity (Casa et al. 2008), which serves as an important genetic resource for the sorghum research community. Interested researchers can use this association panel and phenotype for their trait of interest without the need for further genotyping since the genotypic data along with appropriate statistical models are available for ready use.

A mini core comprising 242 accessions was developed from a core collection of 2,247 accessions through hierarchical cluster analysis using the phenotypic distances estimated from 11 qualitative and 10 quantitative traits and selecting about 10 % or a minimum of one accession per cluster covering a total of 21 clusters. Statistical comparisons based on homogeneity of distribution for geographical origin, biological races, qualitative traits, means, variances, phenotypic diversity indices, and phenotypic correlations indicated that the mini core collection represented the core collection (Upadhyaya et al. 2009). In addition to this, a reference set was developed comprising 374 sorghum accessions through the Generation Challenge Program as a means to enhance utilization of genetic resources in crop improvement (http://www.icrisat.org/what-we-do/crops/sorghum/Sorghum_Reference.htm).

A sorghum diversity research set (SDRS) comprising 107 sorghum accessions representing geographically diverse accessions from 27 countries in Asia and Africa was developed by Shehzad et al. (2009) through the analysis of the genetic diversity of 320 sorghum germplasm accessions with a set of 38 selected SSR markers based on three different published SSR linkage maps of sorghum. A sweet sorghum panel (SSP) was assembled by Murray et al. (2009) comprising 125 diverse accessions, which are primarily old and modern sweet sorghum cultivars along with a few grain and forage sorghums.

A core collection composed of 195 sorghum accessions originating from 39 countries and belonging to the five basic and ten intermediate races representative of the genetic diversity of core sample of 210 cultivated sorghum genotypes reported by Deu et al. (2006) was used for the association mapping of grain quality such as amylose content, protein content, lipid content, hardness, endosperm texture, and peak gelatinization temperature along with grain yield (de Alencar Figueiredo et al. 2010). In addition to the landrace collection described by Deu et al. (2006), an additional 45 inbred lines including donors for aluminum tolerance (Caniato et al. 2007) were used for the association mapping for aluminum tolerance to gain insights into the origin and evolution of aluminum tolerance and to detect functional variants (Caniato et al. 2014). Analysis of recombinant haplotypes suggested that causative polymorphisms are in introns and a transposon (MITE) insertion localized to a ~6 kb region, which is positively correlated with aluminum tolerance. However, the SNP located in the second intron of SbMATE, an Al-activated root citrate efflux transporter, exhibited the strongest association signal and recovered 80 % of all the aluminum-tolerant accessions in the association panel.

7 Mutant Resources for Analyzing Gene Function

Exploitation of natural or induced genetic variability is considered as a successful strategy in crop improvement in many food crops. Mutagenesis as a tool to create novel variation is particularly important for those crops with limited variability. Over the years, several varieties have been developed in major food crops through mutation breeding programs. Ever since H.J. Muller reported induced mutation in 1937, analysis of mutants remained an effective approach to understand the gene function (Springer 2000; Stanford et al. 2001). The rapid accumulation of genomic sequence information in the past decade has brought the reverse genetic approaches into prominence, thereby directly probing the function of specific genes by testing the in vivo consequence of disruption or over-expression of a gene on the phenotype of an organism (Tierney and Lamour 2005). This is the “reverse” of conventional approach where phenotypes are observed and then the gene responsible for that phenotype is cloned and validated.

Mutant lines are important bioresources in this regard, which can potentially accelerate the understanding of gene function through reverse genetics. Kuromori et al. (2009) while reviewing the available mutant resources for phenome analysis in plant species highlighted the importance of mutant bioresource across crop species. With the availability of various analytical platforms (particularly bioinformatics), it is feasible to discover genes involved in particular phenotypic changes. Logically, these genes need to be functionally tested in collection of mutant resources in a high-throughput manner, called phenome analysis (Alonso and Ecker 2006). Chemical mutagens, like ethyl methanesulfonate (EMS), sodium azide, and methylnitrosourea (MNU), and physical mutagens, like fast neutrons, gamma rays, and ion beam irradiation, have been used extensively to generate mutant populations since the report of the first induced mutation in 1937. However, none of these mutant populations have been systematically annotated and preserved (Sree-Ramulu 1970; Porter et al. 1978; Jenks et al. 1994). Thus, the generated resource could not be combined with genomics tools.

Targeting induced local lesions in genomes (TILLING) was developed as a general reverse genetic tool to derive an allelic series of induced point mutations in genes of interest (Till et al. 2004, 2006). TILLING allows rapid and low-cost discovery of induced point mutations in populations of chemically mutagenized individuals in a high-throughput manner. This has been utilized to identify mutations in genes of interest in different crops (Till et al. 2004, 2007; Talame et al. 2008; Lababidi et al. 2009) including sorghum (Xin et al. 2008, 2009; Blomstedt et al. 2012). TILLING resources have been created for various crop plants across laboratories (Barkley and Wang 2008). Xin et al. (2008) were first to report the creation of TILLING resource in sorghum through EMS mutagenesis of sorghum cultivar, BTx623. They documented the feasibility of this approach by screening the mutant population for alterations in the genes of agronomic value not associated with cyanogenesis. Recently, Blomstedt et al. (2012) developed an acyanogenic forage line, P414L, with a point mutation in the CYP79A1 gene of cyanogenesis biochemical pathway by combining biochemical screen and TILLING approach. In the recent years, several TILLING populations have been developed globally by different sorghum research groups, which can serve as a valuable bioresource useful in understanding gene function and high-throughput SNP discovery.

An Annotated Individually pedigreed Mutated Sorghum (AIMS) library comprising 6144 pedigreed M4 seed pools developed through single-seed descent from individual mutagenized seeds was established (Xin et al. 2008, 2009), which contains many biologically and agronomically important mutants, such as brown midrib (bmr) mutants for improved biomass digestibility and ethanol yield and erect leaf (erl) mutants for improved capture of canopy radiation and hence biomass yield (Xin et al. 2009; Saballos et al. 2012; Sattler et al. 2012). An array of useful mutations harboring a wide range of phenotypic variation, including dwarfness, earliness, high protein digestibility, high lysine, etc., that were reported earlier in sorghum (Singh and Axtell 1973; Quinby 1975; Ejeta and Axtell 1985; Oria et al. 2000) were collected and preserved by the late Dr. Keith Schertz, a former sorghum geneticist with USDA-ARS, and this was released recently as a collection of genetic stocks through Germplasm Resources Information Network (Xin et al. 2013; www.ars-grin.gov), which is a valuable and vital resource for future genomic studies in sorghum.

8 Future Prospects

In the current era of crop improvement involving efficient integration of genetic information with the genomic and bioinformatics resources, focus should be on the coordination of various sorghum research groups across the globe on the sharing and utilization of resources available in the public domain. Genomics offers practical advantages for breeding cultivars by providing access to genetic variation through molecular markers and the potential to accurately measure the gene expression. With the initiation of re-sequencing projects in sorghum, whole-genome sequence information of sorghum cultivars possessing different end uses will be available in the near future resulting in the possibility of providing “genotype genomics” services that will contribute to sorghum improvement through “breeding by design.” There is a need for the development of a single platform for sorghum, which can integrate the data that are scattered in different databases and also the software tools available for analysis. The main challenges facing the bioinformaticians are the development and management of databases and computational tools for data analysis in such a way that the user can define the target data and select the computational tool in order to get the output in a suitable format. The availability of different platforms for sequencing demands the development of novel algorithms and computational tools for an efficient assembly, annotation, and analysis. The application of molecular markers, comparative genomics, and annotation tools will greatly assist the identification of the genetic variation underlying an increasing number of agronomic traits and assist in the further agronomic improvement of a variety of crops.