Philosophiae Doctor - PhD / In this dissertation three computational approaches are presented that enable optimization of reference-free transcriptome reconstruction. The first addresses the selection of bona fide reconstructed transcribed fragments (transfrags) from de novo transcriptome assemblies and annotation with a multiple domain co-occurrence framework. We showed that selected transfrags are functionally relevant and represented over 94% of the information derived from annotation by transference. The second approach relates to quality score based RNA-seq sub-sampling and the description of a novel sequence similarity-derived metric for quality assessment of de novo transcriptome assemblies. A detail systematic analysis of the side effects induced by quality score based trimming and or filtering on artefact removal and transcriptome quality is describe. Aggressive trimming produced incomplete reconstructed and missing transfrags. This approach was applied in generating an optimal transcriptome assembly for a South African isolate of V. inaequalis. The third approach deals with the computational partitioning of transfrags assembled from RNA-Seq of mixed host and pathogen reads. We used this strategy to correct a publicly available transcriptome assembly for V. inaequalis (Indian isolate). We binned 50% of the latter to Apple transfrags and identified putative immunity transcript models. Comparative transcriptomic analysis between fungi transfrags from the Indian and South African isolates reveal effectors or transcripts that may be expressed in planta upon morphogenic differentiation. These studies have successfully identified V. inaequalis specific transfrags that can facilitate gene discovery. The unique access to an in-house draft genome assembly allowed us to provide preliminary description of genes that are implicated in pathogenesis. Gene prediction with bona fide transfrags produced 11,692 protein-coding genes. We identified two hydrophobin-like genes and six accessory genes of the melanin biosynthetic pathway that are implicated in the invasive action of the appressorium. The cazyome reveals an impressive repertoire of carbohydrate degrading enzymes and carbohydrate-binding modules amongst which are six polysaccharide lyases, and the largest number of carbohydrate esterases (twenty-eight) known in any fungus sequenced to date

Massively parallel whole transcriptome sequencing and its ability to generate full transcriptome data at the single transcript level provides a powerful tool with multiple interrelated applications, including transcriptome reconstruction, gene/isoform expression estimation, also known as transcriptome quantification. As a result, whole transcriptome sequencing has become the technology of choice for performing transcriptome analysis, rapidly replacing array-based technologies. The most commonly used transcriptome sequencing protocol, referred to as RNA-Seq, generates short (single or paired) sequencing tags from the ends of randomly generated cDNA fragments. RNA-Seq protocol reduces the sequencing cost and significantly increases data throughput, but is computationally challenging to reconstruct full-length transcripts and accurately estimate their abundances across all cell types. We focus on two main problems in transcriptome data analysis, namely, transcriptome reconstruction and quantification. Transcriptome reconstruction, also referred to as novel isoform discovery, is the problem of reconstructing the transcript sequences from the sequencing data. Reconstruction can be done de novo or it can be assisted by existing genome and transcriptome annotations. Transcriptome quantification refers to the problem of estimating the expression level of each transcript. We present a genome-guided and annotation-guided transcriptome reconstruction methods as well as methods for transcript and gene expression level estimation. Empirical results on both synthetic and real RNA-seq datasets show that the proposed methods improve transcriptome quantification and reconstruction accuracy compared to previous methods.