An Automatic Bioinformatics Tool to Detect and Discover Plant Viruses

The speed of plant virus discovery has been further accelerated by the emergence of high throughput sequencing techniques. As the large volume of pyrosequencing and de novo synthesis sequence data has made processing a burden for sceintists, we have designed bioinformatics pipelines to analyze the data automatically. Different samples were extracted, reverse-transcribed and PCR-amplified with up to 48 barcodes (expandable) before sequencing. The pipelines, written in Perl and Shell scripts and making use of NCBI Blast tools, Fastx-Toolkit  and an ORF finder, were able to identify barcodes, sort the data to original samples, clip adapters, filter sequence quality, collapse identical sequences, detect known virus/viroid, and remove plant sequences using Blastn and subsequently Blastx. Sequences without similarity to known organisms were subject to a Python script to identify potential ORFs. The pipelines were tested successfully and independently by users with beginner bioinformatics knowledge to detect viruses from virus-infected samples extracted by total/double-stranded/small RNA extraction methods.