![]() Misjoin editor threshold repeat coverage (default is 2).įor juicebox, you’ll need two files from 3d-dna, your final. editor-repeat-coverage editor_repeat_coverage editor-fine-resolution editor_fine_resiolution Misjoin editor saturation parameter (default is 5). editor-saturation-centile editor_saturation_centile Misjoin editor stringency parameter (default is 55). editor-coarse-stringency editor_coarse_stringency ![]() ![]() Misjoin editor triangular motif region size (default is 125000). editor-coarse-region editor_coarse_region editor-coarse-resolution editor_coarse_resolution ![]() q|-mapq mapq Mapq threshold for scaffolding and visualization (default is 1). If changing normalization doesn’t work, then the “ADDITIONAL OPTIONS” in need to be modified until there is enough coverage for 3d-dna to operate. Vi /02_WhiteMysteryPseudomolecule/04_3ddnaRepCov20/3d-dna/split/run-asm-splitter.sh Modify the normalization for your HiC reads at this line. At high resolution, we suggest VC as the alternative to KR if KR doesn't work for some reason at low resolution, square root VC better approximates KR" With square root, this effect is dampened, but the best remains KR. Places where coverage was high in the original map are "overcompensated" and become depleted. "One way to see the overcorrection of vanilla coverage is to look at supplemental figure S1b. If you see an error that is something like this, "Unknown resolution: BP_100000", then you need to modify the resolution that 3d-dna uses for this step or modify the normalization methods, which are described in the Aiden forum. Create conda environment for installation LastZ installation is not necessary, as it is only useful for highly heterozygous genomes. When I install I usually use singularity containers or conda evironments to address my lack of sudo privileges. Set up and run 3D-DNA assembly Install 3d-dna If juicer does not complete, you can restart from the last completed stage using -S with one of the following options “chimeric”, “merge”, “dedup”, “final”, “postproc”, or “early”. You will need to go into the debug folder, which will be /02_WhiteMysteryPseudomolecule/01_JuicerSetup/debug for me. In most cases this will be a memory issue that can be fixed by creating a greater number of splits of your fastq files. If you have a merged_sort.txt file, but lack the merged_nodups.txt, then the read deduplication step failed. If you want to know why they failed, then you’ll have to investigate the debug folder. If you do not have a merged_sort.txt, the first thing to check is if your split fastq files (in splits/) each aligned and created a sam file. Note that these files are usually hundreds of MB. If at the end of your run, you are seeing large merged_sort.txt and a merged_nodups.txt files, then you likely succeeded. Your stdout and stderr will tell you almost nothing about your run. Ml jdk/10.0.2_13-fr57jru ml juicer juicer.sh -d 01_JuicerSetup -p chrom_sizes -s none -z 01_JuicerSetup/references/ -q short -Q 2:00:00 -l medium -L 12:00:00 -t 36 #/02_WhiteMysteryPseudomolecule/01_JuicerSetup ![]() Softlink and index the genome sequence in the references/ folder The aiden lab forum is an invaluable resource. If you have problems with installation and/or need help with understanding a result. There is lots of information to be had here at the juicer github site. It is not always clear where to find the information you need to run the pipeline, so I will lay out the many resources that I accessed while learning to run this pipeline. Gather the necessary files, setup the juicer file structure, and run juicer Resources With restriction enzyme generated HiC, you will have to create a restriction_sites/ folder and run your genome you will have to creat the proposed restriction enzyme cut sites for your genome. The steps are a bit different, as with an endonuclease you can just say “-s none”. If you plan to use the HiC for anything further than HiC, you will want to specify this. I do not know if this is something that affects the downstream product, but it is an important consideration, considering that long read assemblies are typically error prone.Īnother thing to consider is the type of HiC you are getting, we’ve used both endonuclease and restriction enzyme generated HiC with success. Typically we assemble the genome with long reads and polish (long or illumina reads) prior to running the pipeline. Here at the ISUGIF, we have used multiple different contig assemblers and had success with this pipeline. Genome scaffolding with the Juicer, JuiceBox, 3D-DNA pipeline ![]()
0 Comments
Leave a Reply. |