Pubblicazione scientifica

Microarray-Based quality assessment as a supporting criterion for de novo transcriptome assembly selection

FILONE TECNOLOGICO	TEMA
produzione e raccolta

RIASSUNTO

RNA-Sequencing and de novo assembly have enabled the analysis of species with non-available reference transcriptomes, although intrinsic features (biological and technical) induce errors in the reconstruction. A strategy to resolve these errors consists of varying assembling process parameters to generate multiple reconstructions. However, the best assembly selection remains a challenge. Quantitative metrics for quality assessment have been inconsistent when compared with pertinent references. In this paper, a criterion for supporting assembly selection based on mapping DNA microarray hybridized probes to assembly sets is proposed. Mouse and fruit fly RNA-Seq datasets were assembled with standard de novo procedures. Quality assessment was estimated using quantitative metrics and the proposed criterion. The assembly that best mapped to the available reference transcriptomes of these model species provided the highest quality assembly. The hybridized probes identified the best assemblies, whereas quantitative metrics remained inconsistent. For example, subtle probe mapping difference of 0.25 %, but statistically significant (ANOVA, p $< 0.05$), enabled the assembly selection that led to identify 3,719 more contigs and led to 1,049 further mapped contigs to the mouse reference transcriptome. The microarray data availability for non-model species makes the proposed criterion suitable for quality assessment of multiple de novo assembly strategies.

DATA

Data di pubblicazione: 27/07/2018

AUTORI

PATRICIA CARVAJAL-LOPEZ
FERNANDO D VON BORSTEL
AMADA TORRES
GABRIELLA RUSTICI
JOAQUIN GUTIERREZ
EDUARDO ROMERO-VIVAS

ENTE DI AFFERENZA

GRUPO DE INVESTIGACION EN BIOINFORMATICA CTR INVESTIGACIONES BIOLOGICAS DEL NOROESTE SC LA PAZ BAJA CALIFORNIA SUR MEXICO E MAIL PCARVAJAL PG CIBNOR MX
GRUPO DE INVESTIGACION EN BIOINFORMATICA CTR INVESTIGACIONES BIOLOGICAS DEL NOROESTE SC LA PAZ BAJA CALIFORNIA SUR MEXICO E MAIL FBORSTEL CIBNOR MX
GRUPO DE INVESTIGACION EN BIOINFORMATICA CTR INVESTIGACIONES BIOLOGICAS DEL NOROESTE SC LA PAZ BAJA CALIFORNIA SUR MEXICO E MAIL AMADA TORRES GMAIL COM
GENETICS DEPT UNIV CAMBRIDGE CAMBRIDGE CB2 3EH UK E MAIL GR231 CAM AC UK
GRUPO DE INVESTIGACION EN BIOINFORMATICA CTR INVESTIGACIONES BIOLOGICAS DEL NOROESTE SC LA PAZ BAJA CALIFORNIA SUR MEXICO E MAIL JOAQUING04 CIBNOR MX
CTR INVESTIGACIONES BIOLOGICAS DEL NOROESTE 42579 LA PAZ BAJA CALIFORNIA SUR MEXICO 23090 E MAIL EVIVAS CIBNOR MX

RIVISTA

IEEE/ACM Transactions on Computational Biology and Bioinformatics (Volume: PP, Issue: 99, Page(s): 1)

indietro