GitHub - elswob/SCUBAT: Scaffolding Contigs Using Blat And Transcripts (original) (raw)

#Overview

SCUBAT (Scaffolding Contigs Using BLAT And Transcripts) uses any set of transcripts to identify cases where a transcript is split over multiple genome fragments and attempts to use this information to scaffold the genome.

#Software Requirements

CAP3 - to assemble scaffolds

GNU parallel - to run the CAP3 assemblies in parallel

#Data requirments

1. A genome in FASTA format
2. A set of transcripts in FASTA format
3. A BLAT psl file aligning the transcripts to the genome

#Quick guide

scubat.pl -t genome.fa -q transcripts.fa -p blat_output.psl

To see all options just run scubat.pl

#Details

SCUBAT procedes via 6 steps:

###1) Align the transcripts to the genome (not part of the script):

• Externally - Use BLAT to align the transcripts to the genome with psl output, e.g. blat -t=dna genome.fa -q=dna transcripts.fa -noHead out.psl

###2) Identify informative split transcripts:

• All transcripts that mapped to more than one contig are flagged. Each mapping is then analysed to identify those that map consecutive non-overlapping sections of the transcript on separate contigs allowing for an overlap buffer zone of up to 20bp.

###3) Create scaffolds:

• Each transcript-contig complex is assembled by orientating the contigs based on the BLAT information and adding 10 N’s in-between the contigs.

###4) Cluster scaffolds into groups and assemble:

• The contigs used in each transcript-contig complex are then cross-referenced and any complexes sharing a contig are grouped. Groups are then assembled using CAP3 with default parameters.

###5) Filter the assemblies:

• The assemblies are parsed to identify successful assembly events. Failed assemblies are identified and the largest complex kept, the remainder removed.

###6) Create new contig set:

• A new contig set is generated using the successful CAP3 assemblies, the remaining complexes and any contig not involved.