Functional Annotation and Transcript Expression Analysis of RNA-Seq Data (via NGS) from White Kelampayan (Neolamarckia cadamba) using bioinformatics approach

• Lim Leong Rui

Table of Contents

LIST OF ABBREVIATIONS

LIST OF TABLES

Summary

Neolamarckia cadamba or white kelampayan is a fast growing tree species that generates economic benefits in 8 to 10 years. There are two objectives proposed in the study: to define functional annotation of RNA-Seq data from white kelampayan; to identify and analyse transcript expression level of the RNA-Seq data from the kelampayan tree. The preparation before analysis such as sample collection, RNA isolation, cDNA library construction, sequencing and mapping are done. Later, analysis steps including Blast2Go program application, WEGO application, statistics on proportions test and corrected p-value are mentioned. Functional annotation of RNA-Seq data from white kelampayan is expected to be defined and described in a bar or pie charts format. Meanwhile, the transcript expression level of each sample data of the xylem and leaves tissue is expected to be measured and compared to each other respectively. High-throughput results from the experiment give high genome data quality of the kelampayan tree.

1.0 Introduction

White kelampayan or Neolamarckia cadamba is a fast growing tree species that generates economical benefits to the society in 8 to 10 years (Ho et al., 2014). The tree species plays its important role in pulp and paper production, medical industry, plywood industry as well as furniture production (Joker, 2000; Ho et al., 2014). Since the kelampayan tree gives great economical benefits, many timber companies and research institutes start doing molecular researches on the kelampayan tree species to find out the future potential of the kelampayan. However, less genetic information researches have been generated on the kelampayan tree.

Although certain plants such as Saccharomyces cerevisae and Arabidopsis thaliana were investigated and studied via using next generation sequencing (NGS) technologies (Wang et al., 2010), there was no any NGS research on the kelampayan tree species. NGS is exposed commercially in 2005 (Bubnoff, 2008) and is an alternative way to solve the limitations faced by Sanger sequencing such as high cost, time constrain and complex use. There are several types and applications of NGS widely used nowadays. NGS technologies such as 454 sequencing technology, Illumina sequencing and ABI/SOLiD sequencing system are applied among the researchers in general, while applications of NGS are RNA-Sequencing (RNA-Seq), genomic sequencing and epigenetic applications (Morozova & Marra, 2008; Perdacher, 2011).

In this study, RNA-Seq data via Illumina sequencing will be applied to investigate the transcript expression level of white kelampayan tree species. The transcript expression

level will be calculated and compared statistically (Zheng et al., 2012). Other than that, the functional annotation of the white kelampayan species will be determined and categorized in three parts: biological process, molecular function and cellular component by applying gene ontology (GO) via Blast2GO program (Blast2GO^® – Software for biologists, 2011; Xiong, 2006).

Therefore, the objectives of the study are:

1. To define the functional annotation of RNA-Seq data from white kelampayan in term of gene ontology via Blast2GO program.
2. To determine and analyse transcript expression level of RNA-Seq data from white kelampayan.

2. Literature Review
   1. White kelampayan

White kelampayan or scientifically known as Neolamarckia cadamba, is a fast growing tree species. It is widely distributed in some East Asia countries such as India, Thailand, Malaysia and Papua New Guinea (Joker, 2000). The taxonomy of the tree species is shown as the following (Dubey et al., 2011):

Table 1.1 Taxonomy of white kelampayan tree species

Kelampayan is normally applied for restoration in weather-beaten areas (Joker, 2000). Besides, both leaves and bark of the kelampayan play important role in medical world. The leaves are extracted to serve as mouth wash, while the dried bark is used to relieve fever (World Agroforestry Centre, as cited in Ho et al., 2014). Furthermore, other parts of the kelampayan such as trunk and branches are also used in pulp and paper industry and furniture industry (Joker, 2000). In India, there has some research shown that the flower of the kelampayan can be extracted out to produce essential oil, which can be further produced as Indian perfumes with sandalwood base (Krisnawati et al., 2011).

2. Next Generation Sequencing (NGS)

Next generation sequencing (NGS) is an alternative way to overcome limitations of the first generation sequencing, Sanger sequencing. It was firstly introduced in 2005 (Morozova & Marra, 2008), and this gives a huge impact to the computational biology world.

It brings much advantages compared to Sanger sequencing in term of time efficiency and cost. According to Bubnoff (2008, p.721), he stated that “NGS technology is up to 200 times faster and cheaper than the traditional Sanger sequencing.” He also mentioned that NGS technologies simplify the bacterial cloning process.

There are three types of NGS technologies, which are 454 sequencing technology, Illumina sequencing and ABI/SOLiD sequencing system (Bubnoff, 2008; Morozova & Marra, 2008; Perdacher, 2011). These three technologies have a same feature where the DNA can be amplified via polymerase chain reaction (PCR) without applying any bacterial cloning process (Bubnoff, 2008). Furthermore, there are several applications used in NGS technology such as transcriptome sequencing or RNA-Sequencing (RNA-Seq), genomic sequencing and epigenetic applications which use CHIP-Seq and methylation profiling to work out analysis on interaction in between proteins and DNA and analysis on regulating chromatin structure respectively (Perdacher, 2011).

3. RNA-Sequencing (RNA-Seq)

RNA-Seq, known as Whole Transcriptome Shotgun Sequencing, is “a revolutionary tool for transcriptomes” (Perdacher, 2011; Wang et al., 2010, p. 57). It has been applied in some studied objects such as Saccharomyces cerevisae, Schizosaccharomyces pombe, Arabidopsis thaliana, mouse and human cells (Wang et al., 2010).

RNA-Seq is widely applied in scientific study because it can give a clearer and more understanding image about transcriptomes compared to DNA microarray and serial analysis of gene expression (SAGE) approach. Comparing to these methods, the RNA-Seq is cheaper, more accurate and more time efficient. There are the benefits of RNA-Seq listed as the followings (Getting started with RNA-Seq data analysis, 2011; Nagalakshmi et al., 2010; Perdacher, 2011; Wang et al., 2010) :

1. Species- and transcript-specific probes are not required in the RNA-Seq method and thus, the result data about the nature of the transciptomes is not be easily affected by previous assumption.
2. A hypothesis-free experiment can be designed and created.
3. Species with low resolution of genome annotation can be investigated in a high throughput way.
4. Transcription start codon and boundaries can be easily located and identified, while exon expression and splicing variants can be measured in precise.
5. Undefined genome sequences from non-model organisms such as centipedes can be studied and determined via RNA-Seq.

4. Functional annotation and Gene ontology
   1. Functional annotation

Functional annotation is a term where the information about a gene’s identity such as biological process, cellular component and molecular component is collected, analyzed and described by referring controlled vocabularies, that is gene ontology (GO)(Berardini et al., 2010).

2.3.2 Gene ontology (GO)

GO is a method where the various vocabularies about biological process, cellular component and molecular functions are standardized via consortium of model organisms database (MODs) (Xiong, 2006). He also stated that three parts of GO: biological process, cellular component and molecular functions are described in a hierarchy way, in which the specificity of a functional gene is described from general (top level) to more specified (low level).

4. Reads per kb per million reads (RPKM) and Blast2GO^® program
   1. RPKM

The RPKM is a method where the calculation of gene expression is not influenced by the gene length and sequencing discrepancy (Zheng et al., 2012). The comparison of gene expression between samples can be directly determined once RPKM is used.

2. Blast2GO^® program

Blast2GO, a software tool, was developed in 2005. It was developed to overcome limitations faced in applying gene ontology (GO) terms such as low throughput sequence annotation, low visualization degree and high restriction to annotated sequences from public database. The software tool is initiated by 5 processes: Blast searching from public database such as NCBI, mapping to extract GO terms, application of annotation rule in annotation step, statistical analysis which performs in bar or pie charts and lastly, visualization process. There are many features concerned from the program: vocabularies, data mining, high configuration, high-throughput, user-friendly and low maintenance.

5. WEGO Tool

WEGO, or Web Gene Ontology Annotation, is a useful web tool playing its role in graph plotting, visualization and comparison. By using the WEGO, a histogram with GO annotation results is created via directed acyclic graph (DAG) structure.. According to Ye et al. (2006), the WEGO tool has been widely applied in rice genome project and silkworm genome project. They also mentioned that the web tool is user –friendly and operating system independent, which allows user easy to manipulate the GO annotation distribution graph plotting.

3. Methodology

Developing xylem and leaves tissues were collected from a 2-years old kelampayan tree. RNA was extracted from the collected tissues and further prepared for cDNA library construction (Ho et al., 2014). cDNA libraries were constructed by using ScriptSeq^TM Complete Kit (Epicentre, USA) and Illumina HiSeq 2500 (Illumina Inc. USA) was applied to carry out sequencing. After sequencing, using CLC Genomics Workbench 7.5 (Qiagen, Denmark), low quality reads and unwanted adaptors were removed by quality trimming process with the parameters: quality limit = 0.01 and ambiguous limit = 2. Good quality of the sequencing reads were obtained after trimming.

The sequencing reads were mapped to the reference transcriptome. Using Blast2GO program (BioBam Bioinformatics S.L., Spain), the reference transcriptome was created by assembling the ESTs and transcripts of the kelampayan tree (BioEasy Sdn Bhd, 2014). The sequence similarities were identified by comparing reference transcriptome to NCBI non-reductant protein database. Several parameters were included in blast searching process: expectation value and hit number thresholds (Conesa et al., 2005). Then, the reference transcriptome was mapped to the Gene Ontology terms by referring sequence similarities. During mapping, Blast hit gene identifiers and gene accessions were included in retrieving GO terms for the hit sequences with evidence codes (Conesa et al., 2005). Assembly of a set of candidate annotations and annotations source was done after mapping.

Annotation rule (AR) will be applied to the Gene ontologies of each candidate annotation. Besides, annotation score (AS), including direct term and AT of the annotation rule will be calculated for each candidate annotation (Conesa et al., 2005). Statistics on gene ontology of candidate annotation will be carried on by applying Fischer Exact Test associating with robust false discovery rate (FDR) correction (Blüthgen et al. as cited in Conesa et al., 2005). In addition, a list of significant GO terms will be created via corrected P-values. Bar or pie charts will be used in describing results statistically. The progress in the annotation process for each sequence will be visualized by observing color changes of nodes in the visualization graph. A web tool, WEGO (Ye et al., 2006) will be applied in plotting the distribution of GO annotations results into a histogram.

Using CLC Genomics Workbench 7.5 (Qiagen, Denmark), RNA-Seq analysis will be selected to run in a Batch mode. There will be 2 types of good quality sequencing reads data will be selected before analyzing process. Besides, reference transcriptome will be selected. The parameters for the analysis such as mismatch cost, length fraction and similarity fraction will be set as default. Expression value will be calculated by normalization of the sequencing reads in reads per kilo base per million (RPKM). After that, an experiment on transcript expression level will be set up, where the types of samples selection, experiment type definition and assignation of group names will be considered and included. Then, a track list for the transcript expression level will be created. One of three statistical analyses, “On Proportions”, will be applied in measuring transcript expression level.

On proportions analysis, especially Z test (Kal et al., 1999), four columns (“Proportions difference”, “Fold Change”, “Test statistics” and “P-value”) will be created and added to the experiment table for each analyzed groups pairs: Xylem and leaves tissue groups. The corrected p-values will be set with default calculations: Bonferroni corrected p-values or FDR corrected p-values. Expression analysis of samples will be observed and recorded. Results of statistical analysis on transcript expression level will be visualized in volcano plots with CLC Genomics Workbench 7.5 (Qiagen, Denmark).

4.0 Expected Outcomes

The functional annotation of RNA-seq data from white kelampayan will be expected to be defined in term of gene ontology via Blast2GO program. The transcript expression level of RNA-Seq data from white kelampayan’s xylem and leaves will be expected to be determined and analyzed. High throughput analysis data will be obtained in order to improve genetic data quality and transcriptome identity of kelampayan tree.

5. Work Schedule

6. References

Lim Leong Rui 36728 13