6.3.2.2. unstarch¶
With high-throughput sequencing generating large amounts of genomic data, archiving can be a critical part of an analysis toolkit. BEDOPS includes the unstarch
utility to recover original BED input and whole-file or per-chromosome data attributes from archives created with starch (these can be v1.x or v2.x archives).
The unstarch utility includes large file support on 64-bit operating systems, enabling extraction of more than 2 GB of data (a common restriction on 32-bit systems).
Starch data can be stored with one of two open-source backend compression methods, either bzip2
or gzip
. The unstarch utility will transparently extract data, without the end user needing to specify the backend type.
6.3.2.2.1. Inputs and outputs¶
6.3.2.2.1.2. Output¶
The typical output of unstarch is sorted BED data, which is sent to standard output.
Specifying certain options will instead send archive metadata to standard output, either in text or JSON format, or export whole-file or per-chromosome attributes (also to standard output).
6.3.2.2.2. Requirements¶
The metadata of a Starch v2.x archive must pass an integrity check before unstarch can extract data. Any manual changes to the metadata will cause extraction to fail.
6.3.2.2.3. Usage¶
Use the --help
option to list all options:
unstarch
citation: http://bioinformatics.oxfordjournals.org/content/28/14/1919.abstract
binary version: 2.4.41 (typical) (extracts archive version: 2.2.0 or older)
authors: Alex Reynolds and Shane Neph
USAGE: unstarch [ <chromosome> ] [ --elements |
--elements-max-string-length |
--bases | --bases-uniq |
--has-duplicates | --has-nested | --list |
--list-json | --list-chromosomes |
--archive-timestamp | --note |
--archive-version | --is-starch |
--signature | --verify-signature ]
<starch-file>
Modifiers
--------------------------------------------------------------------------
<chromosome> Optional. Either unarchives chromosome-
specific records from the starch archive
file or restricts action of operator to
chromosome (e.g., chr1, chrY, etc.).
Process Flags
--------------------------------------------------------------------------
--elements Show total element count for archive. If
<chromosome> is specified, the result
shows the element count for the
chromosome.
--elements-max-string-length Show the maximum string length over all
elements in <chromosome>, if specified.
If <chromosome> is not specified, the
maximum string length is shown over all
chromosomes.
--bases,
--bases-uniq Show total and unique base counts,
respectively, for archive. If
<chromosome> is specified, the count is
specific to the chromosome, if available.
--has-duplicate-as-string,
--has-duplicate Show whether there is one or more
duplicate elements in the specified
chromosome, either as a numerical (1/0)
or string (true/false) value. If no
<chromosome> is specified, the value
given indicates if there is one or more
duplicate elements across all chromosome
records.
--has-nested-as-string,
--has-nested Show whether there is one ore more nested
elements in the specified chromosome,
either as a numerical (1/0) or string
(true/false) value. If no <chromosome> is
specified, the value given indicates if
there is one or more nested elements
across all chromosome records.
--list List archive metadata (output is in text
format). If chromosome is specified, the
attributes of the given chromosome are
shown.
--list-json,
--list-json-no-trailing-newline List archive metadata (output is in JSON
format)
--list-chr,
--list-chromosomes List all or specified chromosome in
starch archive (like "bedextract --list-
chr"). If <chromosome> is specified but
is not in the output list, nothing is
returned.
--note Show descriptive note, if available.
--signature Display the Base64-encoded SHA-1 data
integrity signature for specified
<chromosome>, or the signatures of the
metadata and all available chromosomes,
if the <chromosome> is unspecified.
--verify-signature Verify data integrity of specified
<chromosome>, or the integrity of all
available chromosomes, if the
<chromosome> is unspecified.
--archive-timestamp Show archive creation timestamp (ISO 8601
format).
--archive-type Show archive compression type.
--archive-version Show archive version.
--is-starch Test if <starch-file> is a valid archive
and print 0/1 (false/true) to standard
output. Unstarch will also return a non-
zero error code if the input file is not
a valid archive.
--version Show binary version.
--help Show this usage message.
6.3.2.2.3.1. Extraction¶
Specify a specific chromosome to extract data only from that chromosome. This is optional; if a chromosome is not specified, data are extracted from all chromosomes in the archive.
$ unstarch chr12 example.starch
...
6.3.2.2.3.2. Archive attributes¶
Archive attributes are described in greater depth in the Starch specification page. We provide an overview here of the major points.
6.3.2.2.3.2.1. Metadata¶
Use the --list-json
or --list
options to export the archive metadata as a JSON- or table-formatted text string, sent to standard output:
$ unstarch --list-json example.starch
{
"archive": {
"type": "starch",
"customUCSCHeaders": false,
"creationTimestamp": "2014-05-01T14:09:29-0700",
"version": {
"major": 2,
"minor": 2,
"revision": 0
},
"compressionFormat": 0
},
"streams": [
{
"chromosome": "chr1",
"filename": "chr1.pid31740.fiddlehead.regulomecorp.com",
"size": "88330",
"uncompressedLineCount": 10753,
"nonUniqueBaseCount": 549829,
"uniqueBaseCount": 548452,
"duplicateElementExists": false,
"nestedElementExists": false,
"signature": "XtnjojM1LyuMnZI4CIneSzgLI5Q="
"uncompressedLineMaxStringLength": 371
},
...
]
}
The --list-chr
(or --list-chromosomes
) option exports a list of chromosomes stored in the Starch archive.
6.3.2.2.3.2.2. Note¶
Using --note
will export any note stored with the archive, when created.
Tip
One can use starchcat to add a new note to an existing Starch archive.
6.3.2.2.3.2.3. Timestamp¶
The --archive-timestamp
option will report the archive’s creation date and time as an ISO 8601 -formatted string.
6.3.2.2.3.2.4. Compression type¶
The --archive-type
option will report the compression type of the archive, either bzip2
or gzip
:
$ unstarch --archive-type example.starch
unstarch
archive compression type: bzip2
6.3.2.2.3.3. Whole-file or per-chromosome attributes¶
6.3.2.2.3.3.1. Data integrity¶
For a specified chromosome, the --signature
operator reports the very nearly unique “signature” or message digest generated from hashing the extracted, post-transform bytes within the chromosome stream with the SHA-1 hash function, followed with an encoding step with the Base64 scheme to turn it into a human-readable string.
If no chromosome is specified, this operator reports the encoded SHA-1 digests of the archive metadata and the signatures of each chromosome stream.
Signatures can be used to compare chromosome streams between two or more archives.
Further, use of the --verify-signature
option with a chromosome name will compare the signature stored in the metadata (the “expected” signature) with an “observed” value generated from extracting the bytes of the chromosome record and hashing them, and then Base64-encoding the result.
If the observed and expected signatures or digests are identical, this validates or verifies the integrity of the chromosome record. A mismatch would result in a non-zero exit state and suggest potential data corruption and the need for further investigation.
6.3.2.2.3.3.2. Elements¶
The --elements
operator reports the number of BED elements that were compressed into the chromosome stream, if specified. If no chromosome is specified, the sum of elements over all chromosomes is reported.
Tip
This option is equivalent to a wc -l
(line count) operation performed on BED elements that match the given chromosome, but is much, much faster as data are precomputed and stored with the archive, retrieved from the metadata in O(1) time.
The --elements-max-string-length
operator reports the maximum string length of BED elements over the specified chromosome, or the maximum string length over all chromosomes, if no chromosome name is specified.
6.3.2.2.3.3.3. Bases¶
The --bases
and --bases-uniq
flags return the overall and unique base counts for a specified chromosome, or the sum of counts over all chromosomes, if no one chromosome is specified.
6.3.2.2.3.3.4. Duplicate element(s)¶
The --has-duplicate
operator reports whether the chromosome stream contains one or more duplicate elements, printing a 0
if the chromosome does not contain a duplicate element, and a 1
if the chromosome does contain a duplicate.
Note
A duplicate element exists if there are two or more BED elements where the chromosome name and start and stop positions are identical. Id, score, strand and any other optional columns are ignored when determining if a duplicate element is present.
Tip
To get a string value of true
or false
in place of 1
and 0
, use the --has-duplicate-as-string
operator, instead.
Note
If the chromosome name argument to unstarch
is omitted, or set to all
, the --has-duplicate
and --has-duplicate-as-string
operators will return a result for all chromosomes (if any one chromosome has one or more duplicate elements, the return value is 1
or true
, respectively). If the chromosome name is provided and the archive does not contain metadata for the given chromosome, these operators will return a 0
or false
result.
6.3.2.2.3.3.5. Nested element(s)¶
The --has-nested
operator reports whether the chromosome stream contains one or more nested elements, printing a 0
if the chromosome does not contain a nested element, and a 1
if the chromosome does contain a nested element.
Note
The definition of a nested element relies on coordinates and is explained in the documentation for nested elements. Id, score, strand and any other optional columns are ignored when determining if a nested element is present.
Tip
To get a string value of true
or false
in place of 1
and 0
, use the --has-nested-as-string
operator, instead.
Note
If the chromosome name argument to unstarch
is omitted, or set to all
, the --has-nested
and --has-nested-as-string
operators will return a result for all chromosomes (if any one chromosome has one or more nested elements, the return value is 1
or true
, respectively). If the chromosome name is provided and the archive does not contain metadata for the given chromosome, these operators will return a 0
or false
result.
6.3.2.2.4. Example¶
To extract a generic Starch file input to a BED file:
$ unstarch example.starch > example.bed
This creates the sorted file example.bed
, containing BED data from extracting example.starch
. This can be a bzip2
or gzip
-formatted Starch archive—unstarch knows how to extract either type transparently.
To list the chromosomes in a Starch v2 archive, use the --list-chr
(or --list-chromosomes
) option:
$ unstarch --list-chr example.starch
chr1
chr10
chr11
chr11_gl000202_random
chr12
chr13
chr14
chr15
chr16
chr17
...
To show the number of BED elements in chromosome chr13
, use the --elements
operator:
$ unstarch chr13 --elements example.starch
10753
To find the number of unique bases in chromosome chr8
:
$ unstarch chr8 --bases-uniq example.starch
545822
To report if the chromosome chr14
contains at least one duplicate BED element:
$ unstarch chr14 --has-duplicate-as-string example.starch
true
To show when the archive was created:
$ unstarch --archive-timestamp example.starch
2014-05-01T14:09:29-0700
To get the SHA-1 message digest, or “signature” of chromosome chr8
, use the --signature
operator:
$ unstarch chr8 --signature example.starch
nZI4CIneSzgLI5QXtnjojM1LyuM=
The signature is written to the standard output stream.
To verify the data integrity of the same chromosome, use --verify-signature
:
$ unstarch chr8 --verify-signature example.starch
PROGRESS: Expected and observed data integrity signatures match for chromosome [chr8]
Any output from --verify-signature
is written to the standard error stream.
Note
Some option calls will not work with legacy v1.x or v2.0 archives. For instance, to get a result for nested or duplicate elements, you need to input a v2.1 (or greater) archive. If you have a v1.x or v2.0 archive, use the starchcat utility to upgrade an older archive to a Starch v2.2 file, which will recalculate the metadata and make all current attributes available.