TransFlow: a modular framework for assembling and assessing accurate de novo transcriptomes in non-model organisms

Abstract

BackgroundThe advances in high-throughput sequencing technologies are allowing more and more de novo assembling of transcriptomes from many new organisms. Some degree of automation and evaluation is required to warrant reproducibility, repetitivity and the selection of the best possible transcriptome. Workflows and pipelines are becoming an absolute requirement for such a purpose, but the issue of assembling evaluation for de novo transcriptomes in organisms lacking a sequenced genome remains unsolved. An automated, reproducible and flexible framework called TransFlow to accomplish this task is described.ResultsTransFlow with its five independent modules was designed to build different workflows depending on the nature of the original reads. This architecture enables different combinations of Illumina and Roche/454 sequencing data, and can be extended to other sequencing platforms. Its capabilities are illustrated with the selection of reliable plant reference transcriptomes and the assembling six transcriptomes (three case studies for grapevine leaves, olive tree pollen, and chestnut stem, and other three for haustorium, epiphytic structures and their combination for the phytopathogenic fungus Podosphaera xanthii). Arabidopsis and poplar transcriptomes revealed to be the best references. A common result regarding de novo assemblies is that Illumina paired-end reads of 100 nt in length assembled with OASES can provide reliable transcriptomes, while the contribution of longer reads is noticeable only when they complement a set of short, single-reads.ConclusionsTransFlow can handle up to 181 different assembling strategies. Evaluation based on principal component analyses allows its self-adaptation to different sets of reads to provide a suitable transcriptome for each combination of reads and assemblers. As a result, each case study has its own behaviour, prioritises evaluation parameters, and gives an objective and automated way for detecting the best transcriptome within a pool of them. Sequencing data type and quantity (preferably several hundred millions of 2×100 nt or longer), assemblers (OASES for Illumina, MIRA4 and EULER-SR reconciled with CAP3 for Roche/454) and strategy (preferably scaffolding with OASES, and probably merging with Roche/454 when available) arise as the most impacting factors.Electronic supplementary materialThe online version of this article (10.1186/s12859-018-2384-y) contains supplementary material, which is available to authorized users.

Nitric oxide-dependent regulation of sweet pepper fruit ripening

“…The libraries were sequenced and produced a total of 1 204 589 625 raw reads, which were then trimmed, resulting in 1 153 185 684 (95.73%) useful reads (Supplementary Table S1). Owing to the large number of reads, only two replicates from each ripening stage were loaded into TransFlow (Seoane et al , 2018) to obtain the best affordable de novo assembly using these reads. The best transcriptome consisted of 564 642 TTs with a mean length of 791 bp, of which 42.19% were over 500 bp in length (Supplementary Table S2).…”

“…Raw reads were pre-processed, assembled, and evaluated using TransFlow (Seoane et al , 2018), which produces up to 190 different assemblies (including primary and reconciled transcripts) that are then evaluated to find which assembly is closer to the Arabidopsis thaliana and Populus trichocarpa complete transcriptomes. Briefly, reads were pre-processed to remove low-quality, vector, adaptor, low-complexity and contaminant sequences, and organelle DNA, and the undesired segments were trimmed using SeqTrimNext (Falgueras et al , 2010).…”

“…Evaluation of each set of tentative transcripts (TTs) was performed by Principal Component Analysis (PCA) using FactoMineR (Lê et al , 2008). Finally, the set of TTs from the best transcriptome assembly (the one closest to A. thaliana and P. trichocarpa transcriptomes in the PCA) was annotated against the plant division of UniProtKB using Full-LengtherNext (Seoane et al , 2018) with default parameters (minimal identity of ≥45% and minimal TT coverage of ≥25%). This software also allowed annotation with orthologs from the proteomes of A. thaliana (TAIR 10 database, 27 416 proteins), Solanum lycopersicum (ITAG 2.4, 34 725 proteins), Solanum tuberosum (PGSC_DM_v3.4, 52 925 proteins), and C. annuum (bell pepper, UP000222542, 35 548 proteins).…”

The haustorial transcriptome of the cucurbit pathogen Podosphaera xanthii reveals new insights into the biotrophy and pathogenesis of powdery mildew fungi

“…These data are available at NCBI (BioProject PRJNA393391). These reads, along with the 975,070 raw reads previously obtained from a Roche 454 system from hyphal and conidial cDNA, which was generated by oligo dT [25], were used to generate the P. xanthii haustorial and epiphytic transcriptomes using the TransFlow framework [27]. The Module 1 performs the pre-processing of raw Illumina reads (haustorial reads in this study) and their assembly.…”

“…The Illumina raw reads from haustorial RNA and 975,070 Roche 454 raw single-end reads from hyphal and conidial RNA previously published by our laboratory [25] were used to perform the P. xanthii transcriptomic profile. This profile comprises the haustorial transcriptome and the revised version of the epiphytic transcriptome, produced in order to compare it with the haustorial transcriptome, generated using the TransFlow workflow as described in our previous study [27]. TransFlow is a modular framework designed for 454 Roche and Illumina reads that i) pre-processes raw reads, ii) builds several tentative transcriptomes and iii) chooses the best transcriptome.…”