Obtaining a chain file
Here is a non-exhaustive list of chain files (also called LiftOver files). The creators of these chain files are not associated with Genozip in any way, and Genozip does not endorse, recommend or warrant the correctness of any particular file.
​
​
Chain file format
The file format of chain files can be found here: UCSC chain file format.
​
--chain: Chain files and Genozip
Chain files are compressible with Genozip:
genozip GRCh37_to_GRCh38.chain
​
To make the compressed Chain files useable with genozip --chain to lift a VCF file, two reference files are required - the reference genome in Primary coordinates, and the reference genome in Luft coordinates, for example:
​
genozip GRCh37_to_GRCh38.chain --reference hs37d5.ref.genozip --reference GRCh38_full_analysis_set_plus_decoy_hla.ref.genozip
​
The reference files themselves should be prepared from the corresponding FASTA file, for example:
​
genozip --make-reference hs37d5.fa.gz # outputs hs37d5.ref.genozip
​
--match-chrom-to-reference: Fix mismatching contigs
​
It is an unfortunate reality of the current state of bioinformatics, that multiple references exist with differing names for the same contigs. For example, some human references may have a contig called “22” while other references call the same contig “chr22”. Likewise chain files may map between any combinations of these options. Moreover, a chain file may refer to contigs that are absent in the specific reference used.
To address this problem, Genozip provides the option --match-chrom-to-reference that updates contig names to match those in the specified references, and removes alignments of contigs that don’t appear in the references:
genozip GRCh37_to_GRCh38.chain --match-chrom-to-reference --reference hs37d5.ref.genozip --reference GRCh38_full_analysis_set_plus_decoy_hla.ref.genozip
​
--show-chain: Viewing chain file alignments
​
The chain file format is purposely dense, and hence hard to understand visually. To view the alignments contained in a chain file in a human-friendly format, use --show-chain:
​
$ genocat --show-chain GRCh37_to_GRCh38.chain.genozip
##fileformat=GENOZIP-CHAIN
#ALN_I PRIM_CONTIG PRIM_START PRIM_END LUFT_CONTIG LUFT_START LUFT_ENDS XSTRAND ALN_OVERLAP
1 1 10001 177417 chr1 10001 177417 -
2 1 227418 267719 chr1 257667 297968 -
3 1 317720 471368 chr1 347969 501617 X
4 1 521369 1566075 chr1 585989 1630695 -
5 1 1566076 1569784 chr1 1630697 1634405 -
6 1 1569785 1570918 chr1 1634409 1635542 -
7 1 1570919 1570922 chr1 1635547 1635550 -
8 1 1570923 1574299 chr1 1635561 1638937 -
9 1 1574300 1583669 chr1 1638939 1648308 -
​
Column 1 is the sequential number of this alignment.
Columns 2,3,4 are coordinates in the Primary reference genome and columns 5,6,7 are coordinates in the Luft reference genome. Note that these are 1-based coordinates (as in VCF), whereas the Chain file format uses 0-based coordinates.
​
Column 8 contains an X if the range in the Luft reference is reverse complimented.
​
Column 9 contains the ALN_I of an alignment with an overlapping Luft range: This is for cases in which multiple alignments contain overlapping Luft ranges. If this alignment’s Luft range overlaps with the Luft range of more than one other alignment, the column will contain one of them.
​
--show-chain-contigs: Viewing chain file contig information
$ genocat GRCh37_to_GRCh38.chain.genozip --show-chain-contigs
PRIMARY chain file contigs that also exist in the reference file:
PRIMARY 1 length=249250621
PRIMARY 2 length=243199373
PRIMARY 3 length=198022430
PRIMARY 4 length=191154276
PRIMARY 5 length=180915260
LUFT chain file contigs that also exist in the reference file:
LUFT chr1 length=248956422
LUFT chr2 length=242193529
LUFT chr3 length=198295559
LUFT chr4 length=190214555
LUFT chr5 length=181538259
​
Note: All contigs of the respective reference files are shown, even if not referred to in the chain data.
​
Questions? support@genozip.com