Keywords

8.1 Introduction

Finger millet (Eleusine coracana L.) a nutritionally important crop has recently gained a lot of attention due to its nutritional components. The finger millet (FM; also known as ragi) ranks fourth after the Sorghum (Sorghum bicolor), pearl millet (Pennisetum glaucum), and foxtail millet (Setaria italica) worldwide (Kumar et al. 2016). It is an important food crop, which is widely cultivated in arid and semi-arid regions of the world, particularly in eastern Africa; India and the rest of Asia including Sri Lanka and China (Belton and Taylor 2002; Kumar et al. 2016).

Despite being an excellent source of nutrients and minerals with the well-documented health benefits FM, also possess some anti-nutrients (commonly referred to as phytochemicals). The slow digestibility of FM is supplying energy during the day. The plant itself is diaphoretic, diuretic and vermifuge, and its leaf juice was offered to women during childbirth (Antony and Chandra 1998; Vadivoo et al. 1998; Subba Rao and Muralikrishna 2002; Fernandez et al. 2003; Ćujić et al. 2016; Kumar et al. 2018). It has also been used for various diseases, including leprosy, liver disease, pleurisy, pneumonia and smallpox, as a folk remedy. Finger millet has a high fiber content that tends to reduce constipation, high cholesterol, diabetes and intestinal cancer. Besides, it is also suggested that the risk of diabetes and gastrointestinal tract disorders may be effectively minimized with daily intake of FM in diet (Gibson and Helme 2001; Mooser and Carr 2001; Scalbert et al. 2005; Lattimer and Haub 2010; Schatzkin et al. 2007; Fardet et al. 2008).

Despite the fact that many essential nutritional features are available in finger millet grains, research work has largely been ignored to exploit the tremendous potential of this crop for solving the problems associated with protein-energy malnutrition and nutritional deficiencies (Murtaza et al. 2014; Kumar et al. 2018). This may be attributed to the fact that before and during the green revolution, staple crops such as wheat, rice and maize gained a great deal of research attention in order to improve food production. Moreover, the availability of very limited genomic knowledge in past years has further limited finger millet crops improvement programs. However, but, the first time the transcriptome (Rahman et al. 2014; Kumar et al. 2015a, b) and genome data (Hittalmani et al. 2017; Antony et al. 2018) of finger millet is now available due to advances in omics science and technology. After that many papers are now available addressing the complexity of transcriptome data and reported some essential genes/proteins associated with different biological processes and its nutritional potential (Lanham-New 2008; Singh et al. 2014; Chinchole et al. 2017; Kokane et al. 2018; Gupta et al. 2018; Avashthi et al. 2018, 2020; Parvathi et al. 2019). These available data hold enormous information related to millet breeding program. There is a need of time to explore these data and generate new transcriptome data for comparative analysis and mining of novel information through computational and experimental approaches. The chapter highlights the pipeline of finger millet transcriptome analysis from data generation to analysis, integration and deposition of data in public repository for the benefit of the scientific society (Fig. 8.1).

Fig. 8.1
figure 1

Integrated approach for decoding the finger millet transcriptome for novel discovery

Full size image

8.2 Overview of High-Throughput Transcriptome Sequencing Platform

The pyrosequencing-based 454 system by Roche, the sequencing-by-synthesis-based GA/HiSeq/MiSeq devices from Illumina and the sequencing-by-ligation SOLiD system, Pacific Biosciences and Oxford Nanopore are currently the most widely used sequencing platform, and some others are under development (Wolf 2013; Klepikova et al. 2016; Pathak et al. 2018b). These sequencing platforms will be utilized for finger millet transcriptome sequencing as per objective of the research projects (Kumar et al. 2015a, b). The sequencing technology and its estimated experimental cost are mentioned in Table 8.1 (Pathak et al. 2018a, b).

Table 8.1 Sequencing technology and its estimated experimental cost in India; the prices mentioned are for sequencing a single sample with a single library
Full size table

8.3 Experimental Set-Up for Data Generation

As we know that finger millet is one of the agriculturally important crop which holds immense potential for improving human life and prevent from several diseases due to its nutraceuticals properties. Therefore, transcriptome sequencing and its analysis is necessary for identifying key components involved in different biological processes for several purposes. Before, going to the experimental setup for data generation, several points’ need to be addressed which are discussed in the following sections.

8.3.1 Purpose of Transcriptome Sequencing

The starting point for each experiment is to identify its basic aims and determine its viability with respect to the budget and the technique available. For a transcriptome sequencing experiment, questions like, why we need transcriptome data, what type of genes will be characterized or identified and why, what are the future applications of identified genes, etc., need to be addressed.

8.3.2 Statistical Design

It is one of the important steps before going for transcriptome sequencing. Here, we will focus on biological replication and statistical model selection. It is a common practice in many biological investigations. In order to integrate biological replication, the statistical treatment of transcriptome data has moved from single sample studies to more complex statistical designs, such as generalized linear regression models. In case of transcriptome analysis, it will help in sample collection and analysis of data for comparative analysis. For example collections of samples in different time interval as compared to control plants followed by sequencing and analysis. This will help in data analysis for fruitful results.

8.3.3 Choice of Tissue and Time

The abundance of transcript and isoform identity is radically different across tissues, and they change drastically not only throughout embryological development, but also over the entire life of an organism. Therefore, it is important to understand, which tissue and at which, physiological level is most likely to observe difference related to the problem at hand.

8.3.4 Collection of Sample

Sample collections will be done carefully because RNA degrades rapidly. Therefore, after cutting put sample immediately in ice box, generally it a good job for preserving RNA for some time at room temperature, then stored at −80 °C for further utilization and their isolation.

8.3.5 RNA Extraction and Their Quality Assessments

RNA extraction has to be suited to the focal RNA species: during normal mRNA extraction, tiny RNA molecules (<200 bp) will be destroyed following traditional LiCl precipitation or commercially available kits. For small transcripts, such as micro-RNAs, different extraction protocols will be needed. The RNA integrity assessment is a crucial first step for obtaining significant measurements of gene expression.

8.3.6 cDNA Synthesis

Until sequencing, most sequencing platforms usually require RNA to be converted to cDNA. The reverse transcriptase enzymatic reaction may either be prepared by hybridizing an oligo-dT primer onto the mRNA template of poly-A tail or by use of random hexamer primers. It can easily be done by following standard lab protocol or protocol provided through kit.

8.3.7 Library Preparation

Library preparation is an important step and it is platform specific. Library can be prepared as per objective of the research project. During library preparation, cDNA is broken into smaller pieces, which then act as a sequencing template. The cDNA pieces are partly sequenced from one end while a single-end approach is used, where paired-end sequencing read short sequences from both ends. For initial transcriptome assembly and isoform identification, paired-end sequencing may be helpful, but one should be careful that the insert size should not be too big (usually <300 bp), else the small size fraction of transcripts would be lost. On the other hand, too short sizes of insert can give adapter contamination, which may need trimming or removal of reads, leading to complicated analysis.

8.3.8 Sequencing Strategy and Platform Selection

Illumina HiSeq, IonTorrent, Pacific Biosciences are the most commonly used sequencing platform right now and others are under development. For platform selection, cost per base pair, error rate and error profiles, total output and read length are the relevant parameters to note. Where there is a trade-off among read length and total output, for transcriptome data, the latter seems more significant. In de novo assembly, longer reads are very helpful and paired-end reads also work equally well. In the end, what matters is the number of appropriately aligned reads per gene, which defines the precision of the calculation of gene expression.

8.4 Bioinformatics for Data Analysis

Bioinformatics is an interdisciplinary science because it is made from combination of several scientific disciplines, i.e. plant science, animal science, chemical science, physical science, pharmaceutical science, mathematical and statistical science supported by computer science and information technology as support system. We can’t think about transcriptome analysis without bioinformatics. It is playing vital role in dissecting the complexity of transcriptome data for identification of key genes/proteins involved in various biological processes and their expression with respect to time in different tissues. A strong computational skill is required for handling of the transcriptome data and softwares. Several important tools used for transcriptome analysis are highlighted in Table 8.2.

Table 8.2 A list of software used for analysis and integration of high-throughput transcriptome sequencing data along with their application and availability
Full size table

8.4.1 Computational Resources and Programming Skills

During transcriptome sequencing, a big amount of data will be generated. Therefore, we need good computational resources in terms of data storage and analysis, because during data analysis other files such as assembled file, ban, sam, etc., will be generated in the form of results, which also take big space in the computer. Besides, good programming skills are necessary for fetching of key sequences, their annotation and analysis. R, Perl and python are the most demanding language right now in this area. Good command in UNIX/Linux operating systems is must for transcriptome analysis.

8.4.1.1 Understanding of File Format

At the time of transcriptome sequencing and analysis, various files will be generated. The knowledge of these files formats such as fastq, fasta, sam, bam, vcf, etc., is necessary to understand data and results.

8.4.2 Quality Analysis of Generated Data

After transcriptome sequencing, the generated data will be analyzed using bioinformatics tools such as fastqc to determine the quality of generated data. If some error has occurred in data, the software like trimmomatic, cut-adapter, etc., will be run to remove low quality reads from the data. Further, fastqc will be run to evaluate the quality of data after trimming.

8.4.3 Method for Data Assembly

It is a computational method to reconstruct longer sequence (e.g. a transcript) from sort sequence reads. Basically, two methods are available for the assembly of transcriptome data, i.e. de novo assembly and reference-based assembly. In finger millet, the first de novo assembled transcriptome was published in the year 2013.

8.4.3.1 De novo Assembly

De novo assembly uses more computational resources as compared to reference-based assembly. A computer may contain at least 8 cores and 256 GB of RAM to allow assembly within a suitable time frame. An assembled transcript facilitates gene expression studies and annotations for novel discovery. Usually, de novo assembly is more challenging and less accurate than reference-based assembly.

8.4.3.2 Reference Based or Genome-Guided Assembly

Here, the genome of a target organism (if available) or closely related organism will be taken as reference. The transcriptome/RNA-seq reads will be mapped on the reference genome to construct the longer leads from small reads for further investigation. This is considered as more accurate method for assembly of transcriptome data.

8.4.4 Annotation of Assembled Transcript

Annotation is a very common term in Bioinformatics. It is a process to find out biologically important regions in sequences, its expression with respect to particular condition and time, and their involvement/role in different biological processes. The obtained information will be further utilized for validation and other research program.

8.4.4.1 Identification of Differentially Expressed Genes (DEGs)

Identification of differentially expressed genes (DEGs) from transcriptome data is one of the key steps in transcriptome assembly and annotation. If an observed difference or expression level among two experimental conditions is statistically important, a gene is declared as differentially expressed. Bioinformatics analysis of transcriptome data play vital role in investigation of differentially expressed genes. These genes will be further annotated and validated for various uses including crop improvement program through molecular breeding and genetic engineering approaches.

8.4.4.2 Gene Name Assignment

The identified DEGs, i.e. up-regulated and down-regulated sequences will be subjected for gene name assignment based on the available information in databases through computational prediction. Designation of gene name on identified sequences is one of the key steps in characterization of genes and construction of their relationship with known sequences of the related organisms through multiple sequence alignment and building of phylogenetic tree.

8.4.4.3 Gene Set Enrichment Analysis

A set of up-regulated and down-regulated sequences will be subjected to enrichment analysis in terms of gene ontology, i.e. molecular function, biological process and cellular components. Based on gene set enrichment analysis, we will annotate the function of genes identified via transcriptome analysis.

8.4.4.4 Pathway Analysis

Pathway analysis is now being done with bioinformatics applications or web resources that accept and interpret various data from omics (Cirillo et al. 2017). It has become an essential tool for determination of candidate genes involved in different pathways. These studies have promoted an interactive assessment of the genes, their role, regulation or association.

8.4.5 Variant Calling

Development of molecular markers within the putatively functional genomic elements of transcribed DNA is one of the key applications of transcriptomic data. Many tools are available for variant calling, such as the extremely versatile GATK pipeline. Generally, partially overlapping sets of variants will call, as they take various statistical approaches and vary in which parts of the data are used.

8.4.6 Systems Biology for Data Integration and Novel Discovery

Systems biology has emerged in recent years as an effective approach for understanding crop plant systems to improve food production and how their dietary components support our health and avoid diseases, as well as for studying the bioactive molecules that are involved in these impacts (Pathak and Singh 2020). Computational methods and statistical models allow for a broad study of the response of key genes and their role in improving the nutritional content and influence of plant products on human health. This has contributed to the recognition of several essential genes and proteins involved in the growth and nutrition of plants, as well as the discovery of bioactive molecules linked with human, animal and plant health (Kumar et al. 2015a, b, 2018; Pathak et al. 2017a, b; Pathak et al. 2018a; Rana et al. 2020; Pathak and Singh 2020).

8.4.6.1 Pathway Modeling and Simulation Analysis

It is a powerful approach to study the behavior of different genes involved in pathway at different amount/concentration with respect to time. The obtained information from transcriptome analysis and information already available in literatures will be used to build model pathway using systems biology graphical notation (SBGN). The different types of SBGN symbols are available for representing gene, protein, transcription factor, metabolite, etc. (Pathak et al. 2013). We can use these symbols to model pathway. Further, kinetic rate equation will be generated for each species in the model to simulate its dynamic behavior for identification of key components involved in regulation of different molecular mechanisms in the biological systems. A pathway of molecular interactions between calcium exchangers and sensors in different tissues involved in the regulation of calcium transport and their accumulation in seeds of finger millet has been proposed through transcriptomics and systems biology approaches (Fig. 8.2) (Kokane et al. 2018).

Fig. 8.2
figure 2

Figure depicted the molecular mechanism of transport and accumulation of calcium in finger millet seeds

Full size image

8.4.6.2 Network Generation and Analysis

The key genes or proteins obtained from transcriptome analysis will be utilized for network generation and analysis for identification of hubs from the large set of genes. It is a powerful approach that emerged in recent years in the area of biological sciences and made a new discipline called network biology. With the help of network biology approaches, we can sort key genes as hub from large set of genes based on topological parameters for further validation and implementation in the crop improvement program. The modeled pathway will be imported as network via network visualization tools for their analysis. Dehydroascorbate reductase (DHAR) was identified as a key gene regulating different biological processes in Finger millet through network analysis (Fig. 8.3) (Avashthi et al. 2020).

Fig. 8.3
figure 3

Identification of hub genes involved in different biological processes in finger millet through integrated transcriptome data analysis

Full size image

8.4.7 Validation

The results generated from above analysis, i.e. transcriptome sequencing, analysis and their integration with systems biology will be validated through experimental approaches for further implementation in finger millet research program with respect to resistance to abiotic and biotic stresses, food and nutritional security.

8.4.8 Submission of Generated Data in International Data Repository

Advances in sequencing platforms generated huge amount of sequencing data in daily basis right now. Storage and management of these data is a challenging task for bioinformatician. Bioinformatics has tremendous potential in management and analysis of omics data. A lot of database, i.e. primary, composite, secondary, structural and specialized databases are available on internet for management of biological data. Right now Sequence Read Archive, commonly known as SRA (https://www.ncbi.nlm.nih.gov/sra) hosted at National Center for Biotechnology Information (NCBI) is a major resource for submission of transcriptome data. Data related to finger millet transcriptome are also available at SRA for further analysis and their integration with newly sequenced data for novel discovery. A list of finger millet transcriptome sequencing data available in public domains is highlighted in Table 8.3.

Table 8.3 List of finger millet transcriptome sequencing data available in public domain for their further analysis and integration
Full size table

8.5 Application of Transcriptome Sequencing and Data Analysis in Breeding and Improvement of Finger Millet

The availability of finger millet transcriptome sequencing data, re-sequencing approaches and computational resources has led to a new era of breeding, as they make it easier to research the genotype and its relationship to the phenotype, particularly for complex traits. It aids in the discovery of new genes and regulatory sequences, as well as discovery of molecular markers. Besides, breeders will learn about the molecular basis of complex traits via expression studies. They also make it possible to find markers that are related to genes and QTLs. The generated information from the data analysis will be further utilized in the development of smart finger millet crops through molecular breeding or biotechnological approaches for ensuring nutritional security. The identified key candidate’s genes will be also useful in improving nutritional quality, drought resistance, heat resistance, etc., to the other cereal crops, i.e. wheat, rice, maize, etc., via genetic engineering method.

8.6 Conclusion

As per our traditional knowledge finger millet seeds is considered a powerhouse of nutritionally important compound as biochemical factory for ensuring nutritional security. The present chapter highlights the key information related to finger millet transcriptome data analysis for identification of key genes/proteins involved in different biological processes, i.e. calcium accumulation, synthesis of secondary metabolites useful in human health and other information. This will be helpful for the readers for understanding the methodology of transcriptome data analysis and their outcomes, which will be further utilized in crop improvement program for food and nutritional security.