wtsi-hpag / easyChain

easyChain is a pipeline to produce a chain file from two genome assemblies.

easyChain

Pipeline to generate chain file for assembly coordinate conversion.

To perform whole genome alignment between target assembly_1.fasta (GRCh37-assembly) and reference assembly_2.fasta (GRCh38-assembly) following the steps:

       1. The target assembly is shredded into chunks of 20000 bases 
       2. The 20000 bases chunks are mapped against the reference assembly
       3. Generate a standard chain file shredOut.chain using the alignment file 

Download and Compile:

Requirements for compiling: gcc

	$ git clone https://github.com/wtsi-hpag/easyChain.git
	$ cd easyChain 
	$ bash install.sh

If everything compiled saccessfully you must see the final comment: "Congrats: installation successful!"

(Tested with gcc-4.9.2, gcc-4.9.4, gcc-4.8.1, gcc-6.0.2)

External packages

The genome aligner BWA (http://bio-bwa.sourceforge.net) and SMALT (http://www.sanger.ac.uk/science/tools/smalt-0) are downloaded and compiled by easyChain.

Run:

       $ /full/path/to/easyChain/src/easyChain -nodes <nodes> -shred <shred_length> \
   	      </full/path/to/assembly_1.fasta> </full/path/to/assembly_2.fasta> <shredOut.chain>\ 
       
       where:
          /full/path/to/assembly_1.fasta: full path to the assembly file to be considered as "GRCh37 assembly"
     	  /full/path/to/assembly_2.fasta:  full path to the assembly file to be considered as "GRCh38 assembly"
     	  shredOut.chain:   output name for the standard chain file. 
     
       parameters:
         nodes:    number of CPUs requested  [ default = 30 ]
         shred:    length of shredded fragments [ default = 20000 ]
         output:   output file (1) alignment only; (2) standard chain file [ default = 2 ]

Note

 1. The shred2chain part is developed by Yongji Liu in Beijing, China, see

    https://github.com/liu-yongji/shred2chain

    It was written in C++ and some libraries used might be difficut to compile. 
    In this pipeline, I used the pre-complied binary code. 

 2. Please use the fullpath when running easyChain
    /full/path/to/easyChain/src/easyChain

checkError

We provide a pipeline to check conversion errors made by different chain files Before using this tool, you need to have the VCF files ready:

 1. Download and unzip all the VCF files from the 1000 Genome project for GRCH37;( chr1~22，X,Y)
    http://ftp.1000genomes.ebi.ac.uk/vol1/ftp/release/20130502/
    make a directory grch37_vcf and copy all the unzipped GRCh37 VCF files there

 2. Download and unzip all the VCF files from the 1000 Genome project for GRCH38.( chr1~22，X,Y)
    http://ftp.1000genomes.ebi.ac.uk/vol1/ftp/release/20130502/supporting/GRCh38_positions/
    make a directory grch38_vcf and copy all the unzipped GRCh38 VCF files there

 3. Disease databases such as ClinVar, Gwas_catlog, hgmd and omim
    We have included these 4 datasets in the pipeline and they will be ready after installation \ 
    No action is needed for users if you are working on /full/path/to/easyChain/src/            \
    Or copy IMDB to the working directory if you are not working on src                         \ 
 
 4. VCF files have to be the annotated files with RS numbers assigned to each called variant
    Self generated VCFs without annotation will not work

USAGE

[Usage]: ./checkError [VCF_Files_folder] [reference_VCF_Files_folder] [chain_file] [output_folder] \ 
[Example]: ./checkError grch37_vcf grch38_vcf hg19ToHg38.over.chain output_result \ 

grch37_vcf            - The folder which contains some or all the VCF files for GRCh37 \
grch38_vcf            - The folder which contains ALL the VCF files for GRCh38    \
hg19ToHg38.over.chain - The chain file selected                                   \
output_result         - The folder with the output results                        \

Note

You need to create a directory for output results. If not exists, the code will generate a temp one \ Five files will be generated after processing for each input VCF file, they are: \

[1]  xxx_SNP.bed:                   - The bed file extracted from xxx.vcf with tag "VT=SNP".   \
[2]  xxx_SNP_genegos.bed:           - The file after coordinate conversion.                    \
[3]  xxx_SNP_genegos.unmap:         - The content that could not be converted.                 \
[4]  xxx_SNP_genegos_error.dat:     - The file contains all the error sites.                   \
[5]  xxx_SNP_genegos_error_db.txt:  - The file contains all error sites in import databases.   \

Disease databases and error validation

We have collected data from ClinVar, Gwas_catlog, hgmd and omim. After installation of the tool
there is a directory of IMDB under /full/path/to/easyChain/src/. When running the pipeline
the IMDB needs to be in the working directory. Alternatively, download the IMDB fie from

ftp://ftp.sanger.ac.uk/pub/users/zn1/easyChain/IMDB.tar.gz

Error validation is very important for selecting a suitable chain file for use. With the 1000genomes data
the results from Ensembl, UCSC, Genegos and shredOut chain files can be downloaded from

ftp://ftp.sanger.ac.uk/pub/users/zn1/easyChain/output.tar.gz

Further information

 1. The checkError tool is developed by Yongji Liu in Beijing, China, see
    https://github.com/liu-yongji/checkbederror

 2. Zemin Ning integrated all the codes and disease databases here for better download and installation. 

 If you have any problems, please contact

     Zemin Ning ( [e-mail] )