Documentation + Presentation of de novo genome assembly efforts on a 20Mb protist.
See report.html for analysis results. The report was generated using R and knitr.
slides.md contains the presentation held. To create the HTML slides slides.html, you need pandoc and reveal.js (included here with a hacked stylesheet).
All scripts that were used in assembly and report generation are in the subfolder scripts/ or otherwise documented in report.html.
Breakdown:
run_abyss_k64_uncorrected.sh
Assembly with Abyss + Scaffolding using Pacbio reads. Includes workarounds for not being able to use MPI and a crash in the abyss pipeline when running the BWA alignment.
run_cerulean.sh
Call blasr and Cerulean to scaffold Abyss contigs.
run_sga_preqc.sh
SGA quality control for Illumina Libraries.
sga.sh
Run all stages of the SGA assembler. Tries to be intelligent about previously finished stages. Unlike all other scripts, it does not have a run_ prefix for unknown reasons.
run_soap.sh
Run SOAP-denovo with k=81.
runCorrection.sh
pacBioToCA pipeline. Use camel case for script name for unknown reasons.
run_preassembly.sh
start SMRTanalysis pipeline
run_pbjelly.sh
This is only the wrapper script. Please see report.html for details.
run_cegma.sh
Analyses an assembly for presence of core genes (CEGMA pipeline).
align_scaffolds.sh,calc_pileup.pl,error_profile.pl
BWA alignment of PacBio reads to Assembly scaffolds and various analysis scripts (need to run module load perl5lib to set env).
Most data analysis with R and knitr. To rerun the report, go to the data/ subfolder and do the following in R:
library('knitr') knit('../report.Rmd') markdownToBsHTML('report.md', output='report.html')
markdownToBsHTML is a hack to produce nicer results, but includes javascript and stylesheets from Bootstrap (uses markdown.html as a template). This is not necessary, markdownToHTML is enough.
Raw data is not included!