k8hertweck/repipe

characterization of TEs from low-coverage genomic sequence data

REpipe: A workflow for assessing TEs from low-coverage genomic sequence data

• low-coverage, anonymous, next-generation Illumina data using local desktop and Stampede.

Description of files:

• dataTransfer/: reminders for data transfer to/from Stampede
• qc/: taxon and data specific customized scripts for quality control of onion sequences
• masurca.slurm: runs assembly for all taxa in taxa.lst in sequential order
• launchRepipe.slurm: batch file for running REpipeline.slurm for scripts described in REscripts
• troubleshooting/: separate scripts for REpipeline and masurca for different parameter settings (prelim analyses; not necessary to run now)
• table.sh: pulls files from Stampede to local desktop, sets up file structure for parsing results from analyses
• joinplot.R: join individual species files, summarize data, plot results

Workflow

From raw reads:

• FastQC
• Trim and filter
• FastQC

From filtered reads:

• Assemble with brute-force algorithm (MaSuRCA, VELVET, other?)
• Assemble with TE-specific method
• graph-based quantification (RepeatExplorer)

From assemblies (genomes or transcriptomes):

• domainPrep.sh: prepare fasta files for analysis and run RepeatMasker
• RT.sh: runs RPS-BLAST against RTV_1, RTV_2, RT_nLTR_like, extracts fasta, clusters results
• INT.sh: runs RPS-BLAST against rve, extracts fasta, clusters results
• domainTree.sh: align with mafft, build tree wiht RAxML

Domains of interest

• RVT_2 (PF07727) - reverse transcriptase core domain that picks up Copia (and thus Sireviruses as well)
• RVT_1 (PF00078) - reverse transcriptase core domain that picks up Gypsy
• rve (PF00665) - the integrase core domain that works for both Copia/Gypsy